The Elgar Companion To Post Keynesian Economics

THE ELGAR COMPANION TO
POST KEYNESIAN ECONOMICS
The Elgar Companion to

Post Keynesian Economics
Edited by
J.E. King
Professor of Economics, Department of Economics and
Finance, La Trobe University, Melbourne, Australia
Edward Elgar
Cheltenham, UK Northampton, MA, USA
J.E. King 2003

All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system or transmitted in any form or by any means, electronic,
mechanical or photocopying, recording, or otherwise without the prior
permission of the publisher.
Published by
Edward Elgar Publishing Limited
Glensanda House
Montpellier Parade
Cheltenham
Glos GL50 1UA
UK
Edward Elgar Publishing, Inc.
136 West Street
Suite 202
Northampton
Massachusetts 01060
USA
A catalogue record for this book

is available from the British Library
Library of Congress Cataloguing in Publication Data

The Elgar companion to post Keynesian economics / edited by J.E. King.
p. cm.
Includes bibliographical references.
1. Keynesian economics. 1. King, J.E. (John Edward)
HB99.7.E522 2003
330.15dc21
2003044802
ISBN 1 84064 630 6 (cased)

Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall
Contents
viii
ix
x
xiv
List of figures
List of tables
List of contributors
Introduction
Agency Edward J. McKenna and Diane C. Zannoni
Austrian School of Economics Stephen D. Parsons
Babylonian Mode of Thought Sheila C. Dow
Balance-of-payments-constrained Economic Growth
J.S.L. McCombie
Banking Gillian Hewitson
Bastard Keynesianism John Lodewijks
Bretton Woods Matias Vernengo
Budget Deficits Julio Lpez G.
Business Cycles Peter Skott
Cambridge Economic Tradition G.C. Harcourt
Capital Theory Ben Fine
Central Banks Fernando J. Cardim de Carvalho
Circuit Theory Riccardo Realfonzo
Competition Nina Shapiro
Consumer Theory Marc Lavoie
Consumption David Bunting
Credit Rationing Martin H. Wolfson
Critical Realism Andrew Brown
Development Finance Rogrio Studart
Dynamics J. Barkley Rosser, Jr.
Econometrics Paul Downward
Economic Policy Malcolm Sawyer
Eective Demand Mark Setterfield
Employment John Hudson
Endogenous Money Basil Moore
Environmental Economics Adrian Winnett
Equilibrium and Non-equilibrium Donald W. Katzner
Exchange Rates John T. Harvey
Expectations ric Tymoigne
Finance Motive Augusto Graziani
Financial Instability Hypothesis Louis-Philippe Rochon
v
1
5
11
15
20
24
30
34
38
44
51
57
60
65
68
72
77
82
87
92
96
101
105
112
117
122
126
131
135
142
145
vi
Contents
Fiscal Policy J.W. Nevile

Full Employment William Mitchell and Martin Watts
Fundamentalist Keynesians Bill Gerrard
Globalization William Milberg
Growth and Income Distribution Carlo Panico
Growth Theory Steve Keen
Income Distribution Thomas I. Palley
Inflation John Smithin
Innovation Jerry Courvisanos
Institutionalism Steven Pressman
International Economics Robert A. Blecker
Investment Tracy Mott
Joan Robinsons Economics Maria Cristina Marcuzzo
Journal of Post Keynesian Economics Phillip Anthony OHara
Kaldorian Economics A.P. Thirlwall
Kaleckian Economics Jan Toporowski
Keyness General Theory Paul Davidson
Keyness Treatise on Money Giuseppe Fontana
Liquidity Preference Stephanie Bell
Marginalism Harry Bloch
Microfoundations Steven Fazzari
Monetary Policy Peter Howells
Money L. Randall Wray
Multiplier Andrew B. Trigg
New Classical Economics Athol Fitzgibbons
New Keynesian Economics Wendy Cornwall
Non-ergodicity Stephen P. Dunn
Pricing and Prices Frederic S. Lee
Production Amitava Krishna Dutt
Profits Elizabeth Webster
Rate of Interest Massimo Pivetti
Saving Robert Pollin
Says Law Claudio Sardoni
Socialism Howard J. Sherman
Sraan Economics Gary Mongiovi
Stagflation John Cornwall
Taxation Anthony J. Laramie and Douglas Mair
Tax-based Incomes Policy Laurence S. Seidman
Third Way Egon Matzner
Time in Economic Theory John F. Henry
Tobin Tax Philip Arestis
Transition Economies Christine Rider
149
153
159
165
170
175
181
186
191
196
200
205
211
215
221
226
229
237
242
249
252
257
261
265
271
275
280
285
289
294
299
304
309
313
318
322
328
332
337
341
346
350
Contents
vii
Traverse Peter Kriesler

Treatise on Probability Rod ODonnell
Uncertainty Murray Glickman
Underconsumption J.E. King
Unemployment Mathew Forstater
Wages and Labour Markets Mario Seccareccia
Walrasian Economics M.C. Howard
355
359
366
370
374
380
384
Name index
Subject index
389
395
Figures
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
No reswitching
52
Reswitching
53
The monetary circuit
61
The consumption function
74
Wolfsons model of credit rationing
81
The principle of eective demand and labour market outcomes 107
The special case of Says Law
109
The interaction of expected and actual aggregate demand, and
aggregate supply
111
Determination of the level of investment
137
Determination of the level of employment
137
A Says Law economy
231
A Keynesian economy
232
The liquidity preference function
244
Determination of the rate of interest
245
Average annual rates of inflation and standardized unemployment
for the G7 countries, 19671997
324
viii
Tables
1
2
Conceptualizations of economic processes

Saving rates, credit supply and GDP growth for the
US economy
ix
283
305
Arestis, Philip: Levy Economics Institute of Bard College, New York, USA
Bell, Stephanie: University of Missouri, Kansas City, MO, USA
Blecker, Robert A.: American University, Washington, DC, USA
Bloch, Harry: Curtin University, Perth, Australia
Brown, Andrew: University of Leeds, Leeds, UK
Bunting, David: Eastern Washington University, USA
Carvalho, Fernando J. Cardim de: Universidade Federal de Rio de Janeiro,
Rio de Janeiro, Brazil
Cornwall, John: Dalhousie University, Halifax, Nova Scotia, Canada
Cornwall, Wendy: Mount Saint Vincent University, Halifax, Nova Scotia,
Canada
Courvisanos, Jerry: University of Ballarat, Ballarat, Victoria, Australia
Davidson, Paul: University of Tennessee, Knoxville, TN, USA
Dow, Sheila C.: University of Stirling, Stirling, Scotland, UK
Downward, Paul: University of Staordshire, Stoke on Trent, UK
Dunn, Stephen P.: Department of Health, London, UK
Dutt, Amitava Krishna: University of Notre Dame, Notre Dame, IN, USA
Fazzari, Steven: Washington University, St. Louis, MO, USA
Fine, Ben: School of African and Oriental Studies, University of London,
London, UK
Fitzgibbons, Athol: Grith University, Brisbane, Australia
Fontana, Giuseppe: University of Leeds, Leeds, UK
Forstater, Mathew: University of Missouri, Kansas City, MO, USA
Gerrard, Bill: University of Leeds, Leeds, UK
Glickman, Murray: University of East London, London, UK
Graziani, Augusto: University of Rome La Sapienza, Rome, Italy
x
xi
Harcourt, G.C.: Jesus College, Cambridge, UK

Harvey, John T.: Texas Christian University, Forth Worth, TX, USA
Henry, John F.: California State University, Sacramento, CA, USA
Hewitson, Gillian: La Trobe University, Melbourne, Australia
Howard, M.C.: University of Waterloo, Waterloo, Ontario, Canada
Howells, Peter: University of East London, London, UK
Hudson, John: University of Bath, Bath, UK
Katzner, Donald W.: University of Massachusetts, Amherst, MA, USA
Keen, Steve: University of Western Sydney, Sydney, Australia
King, J.E.: La Trobe University, Melbourne, Australia
Kriesler, Peter: University of New South Wales, Sydney, Australia
Laramie, Anthony J.: Merrimack College, North Andover, MA, USA
Lavoie, Marc: University of Ottawa, Ottawa, Canada
Lee, Frederic S.: University of Missouri, Kansas City, MO, USA
Lodewijks, John: University of New South Wales, Sydney, Australia
Lpez G., Julio: Universidad Autonoma de Mexico, Mexico City, Mexico
Mair, Douglas: Heriot-Watt University, Edinburgh, Scotland, UK
Marcuzzo, Maria Cristina: University of Rome La Sapienza, Rome, Italy
Matzner, Egon: University of Technology, Vienna, Austria (retired)
McCombie, J.S.L.: University of Cambridge, Cambridge, UK
McKenna, Edward J.: Connecticut College, New London, CT, USA
Milberg, William: New School University, New York, USA
Mitchell, William: University of Newcastle, Newcastle, Australia
Mongiovi, Gary: St. Johns University, Jamaica, NY, USA
Moore, Basil: Wesleyan University, Middletown, CT, USA and University
of Stellenbosch, Stellenbosch, South Africa
Mott, Tracy: University of Denver, Denver, CO, USA
Nevile, J.W.: University of New South Wales, Sydney, Australia
xii
ODonnell, Rod: Macquarie University, Sydney, Australia

OHara, Phillip Anthony: Curtin University, Perth, Australia
Palley, Thomas I.: Open Society Institute, Washington, DC, USA
Panico, Carlo: University of Naples, Naples, Italy
Parsons, Stephen D.: De Montfort University, Leicester, UK
Pivetti, Massimo: University of Rome La Sapienza, Rome, Italy
Pollin, Robert: University of Massachusetts, Amherst, MA, USA
Pressman, Steven: Monmouth University, West Long Branch, NJ, USA
Realfonzo, Riccardo: University of Sannio, Benevento, Italy
Rider, Christine: St. Johns University, Jamaica, NY, USA
Rochon, Louis-Philippe: Kalamazoo College, Kalamazoo, MI, USA
Rosser, J. Barkley, Jr.: James Madison University, Harrisonburg, VA, USA
Sardoni, Claudio: University of Rome La Sapienza, Rome, Italy
Sawyer, Malcolm: University of Leeds, Leeds, UK
Seccareccia, Mario: University of Ottawa, Ottawa, Canada
Seidman, Laurence S.: University of Delaware, Newark, DE, USA
Setterfield, Mark: Trinity College, Hartford, CT, USA
Shapiro, Nina: St. Peters College, Jersey City, NJ, USA
Sherman, Howard J.: University of California, Los Angeles, CA, USA
Skott, Peter: University of Aarhus, Aarhus, Denmark
Smithin, John: York University, Toronto, Canada
Studart, Rogrio: UNECLAC, Chile
Thirlwall, A.P.: University of Kent, Canterbury, UK
Toporowski, Jan: South Bank University, London, UK
Trigg, Andrew B.: The Open University, Milton Keynes, UK
Tymoigne, ric: University of Missouri, Kansas City, MO, USA
Vernengo, Matias: Kalamazoo College, Kalamazoo, MI, USA
Watts, Martin: University of Newcastle, Newcastle, Australia
xiii
Webster, Elizabeth: University of Melbourne, Melbourne, Australia

Winnett, Adrian: University of Bath, Bath, UK
Wolfson, Martin H.: University of Notre Dame, Notre Dame, IN, USA
Wray, L. Randall: University of Missouri, Kansas City, MO, USA
Zannoni, Diane C.: Trinity College, Hartford, CT, USA
Introduction
Stripped down to the bare essentials, Post Keynesian economics rests on the
principle of eective demand: in capitalist economies, output and employment are normally constrained by aggregate demand, not by individual
supply behaviour. Since a decision not to have lunch today as Keynes
famously put it does not entail a decision to have lunch tomorrow, investment drives saving and not the other way round. Moreover, there exists no
automatic or even minimally reliable mechanism that will eliminate excess
capacity and involuntary unemployment. Interest rates depend on monetary considerations, not on the so-called real forces of productivity and
thrift. There is no natural rate of interest to equilibrate investment
and saving, so that an increase in the propensity to save will prove selfdefeating, resulting in lower output and reduced employment but not in
higher levels of saving.
Thus far Post Keynesians agree with mainstream, neoclassical, old or
less politely put Bastard Keynesians like J.R. Hicks, Paul Krugman, James
Meade and Paul Samuelson. They part company with them, however, in
denying the validity of the neoclassical synthesis and in rejecting the ISLM
model, the real balance eect and the notion of the long run as a sort of
magic kingdom where the future is knowable (at least probabilistically),
expectations are always fulfilled, money has no real significance and all
resources are fully employed. In fact Post Keynesianism emerged as a distinct school of thought, in the 1960s, precisely as a reaction against these
perversions of Keyness original vision. In his General Theory, as they interpreted it, uncertainty was inescapable, expectations were tentative and unreliable, money aected output as well as prices, and demand-deficient
unemployment was the central macroeconomic problem. These issues are
discussed below in the entries on eective demand, employment, Keyness
General Theory, saving, Says Law and unemployment. At roughly the same
time there emerged a thorough and incisive Post Keynesian critique of the
neoclassical theories of capital, growth and distribution, together with an
insistence on the importance of cost inflation and the role of incomes policy
as an indispensable weapon with which to fight it (see the entries on capital
theory, growth and income distribution, growth theory, inflation, stagflation and tax-based incomes policy).
The tendency for Post Keynesians to define themselves through criticism
of the mainstream has led many orthodox economists to conclude that they
have nothing positive to say. This is quite unwarranted, but it does contain an
xiv
Introduction
xv
element of truth: one way of appreciating what Post Keynesians do believe is

through understanding what it is that they reject. The entries on Bastard
Keynesianism, marginalism, New Classical economics, New Keynesian economics and Walrasian economics can profitably be approached from this perspective. A dierent but related objection is that Post Keynesianism is
incoherent when viewed as a set of positive propositions. In an early survey
article Omar Hamouda and Geo Harcourt (1988) identified three, potentially incompatible, streams of Post Keynesian thinking, which they termed
the Fundamentalist Keynesians, the Kaleckians and the Sraans. There are
entries on all three (see fundamentalist Keynesians, Kaleckian economics
and Sraan economics), together with two other heterodox currents often
taken to have something in common with them, the Austrians and the institutionalists (see Austrian school of economics and institutionalism).
Post Keynesian economics is certainly a very broad church, and I have
tried to reflect this diversity in choosing topics and contributors. The entries
on monetary questions provide one example (circuit theory, endogenous
money, finance motive, financial instability hypothesis and money) and
those on policy issues are another (budget deficits, economic policy, fiscal
policy, monetary policy, taxation and tax-based incomes policy). With very
few exceptions, Post Keynesians are hostile to neoliberalism and united in
their support for active macroeconomic management, nationally and internationally (for the latter dimension, see the entries on Bretton Woods,
development finance, globalization, international economics and transition
economies).
The focus of this book is predominantly macroeconomic, though Post
Keynesians have made important contributions to microeconomic theory
and policy (as demonstrated in the entries on agency, competition, consumer theory, environmental economics, and pricing and prices) and on questions of economic philosophy, methodology and research methods (see the
entries on Babylonian mode of thought, critical realism, econometrics,
non-ergodicity and socialism).
Biographical and autobiographical accounts of many prominent Post
Keynesians are readily available elsewhere see especially Arestis and
Sawyer (2000) and thus with only three exceptions there are no biographical entries here (Joan Robinsons economics, Kaldorian economics and
Kaleckian economics). John Maynard Keynes, though, is represented by
entries dealing with his three great books (General Theory, Treatise on
Money and Treatise on Probability), and his presence throughout the
volume is so pervasive that it seemed pointless to provide an entry for him
in the name index.
Each entry contains references to the relevant literature. Readers looking
for an introductory overview of Post Keynesian economics should begin
xvi
Introduction
with the entry on the Journal of Post Keynesian Economics in the present
volume and then proceed to Holt and Pressman (2001), perhaps in conjunction with the book that it replaced (Eichner 1979). A number of survey articles have appeared, beginning with the previously mentioned paper by
Hamouda and Harcourt (1988), and continuing with Arestis (1996) and
Arestis and Sawyer (1998), the latter concentrating on policy. Several of the
essays in Harcourt (2001) will also be useful. Post Keynesian textbooks
include Arestis (1992), Davidson (1994) and at a more advanced level
Lavoie (1992). A history of Post Keynesian ideas is provided by King
(2002), which concentrates on macroeconomics and should be complemented by Lee (1998) on the microeconomic aspects. King (1995) oers a
reasonably complete bibliography up to 1994.
I am grateful to Edward Elgar for suggesting this project to me, and
indeed for his consistent support for Post Keynesian economics over almost
two decades. Philip Arestis and Malcolm Sawyer were extremely helpful at
the start, and Fred Lees assistance was invaluable later on. I must also
thank the contributors for tolerating my sometimes savage editorial
assaults on their early drafts (which in one case amounted to a 75 per cent
cut). It is invidious to single out individuals, but ric Tymoigne does
deserve a special mention for writing lucidly at exceptionally short notice.
Subject to the usual disclaimer, Phillip OHara wishes to thank Harry
Bloch, John King, Peter Kriesler, Marc Lavoie and Douglas Vickers for
comments on his entry, and I benefited considerably from the criticism of
Marc Lavoie and Michael Schneider on my own.
This volume is dedicated to the memory of Bernard Corry, who sadly
died before he could complete the entry that he was working on.
References
Arestis, P. (1992), The Post Keynesian Approach to Economics: An Alternative Analysis of
Economic Theory and Policy, Aldershot, UK and Brookfield, VT, USA: Edward Elgar.
Arestis, P. (1996), Post-Keynesian economics: towards coherence, Cambridge Journal of
Economics, 20 (1), 11135.
Arestis, P. and M. Sawyer (1998), Keynesian economic policies for the new millennium,
Economic Journal, 108 (446), 18195.
Arestis, P. and M. Sawyer (2000), A Biographical Dictionary of Dissenting Economists, 2nd
edition, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Davidson, P. (1994), Post Keynesian Macroeconomic Theory: A Foundation For Successful
Economic Policies For the Twenty-First Century, Aldershot, UK and Brookfield, VT, USA:
Edward Elgar.
Eichner, A.S. (ed.) (1979), A Guide to Post Keynesian Economics, London: Macmillan.
Hamouda, O.F. and G.C. Harcourt (1988), Post-Keynesianism: from criticism to coherence?,
Bulletin of Economic Research, 40 (1), 133.
Harcourt, G.C. (2001), Fifty Years a Keynesian and Other Essays, Basingstoke and New York:
Palgrave.
Holt, R.P.F. and S. Pressman (eds) (2001), A New Guide to Post Keynesian Economics, London
and New York: Routledge.
Introduction
xvii
King, J.E. (1995), Post Keynesian Economics: An Annotated Bibliography, Aldershot, UK and
Brookfield, VT, USA: Edward Elgar.
King, J.E. (2002), A History of Post Keynesian Economics Since 1936, Cheltenham, UK and
Northampton, MA, USA: Edward Elgar.
Lavoie, M. (1992), Foundations of Post Keynesian Economic Analysis, Aldershot, UK and
Brookfield, VT, USA: Edward Elgar.
Lee, F.S. (1998), Post Keynesian Price Theory, Cambridge: Cambridge University Press.
Agency
Agents are the sources of choices and decisions. Agency deals with the
capacity that enables choices to be made. The provenance of this capacity,
its nature, and the factors that enhance or limit it are the main questions of
concern. From the point of view of Post Keynesian economics, an additional question arises. As is well known, an essential aspect of Post
Keynesian economics is the adoption of the non-ergodicity postulate,
which states that probability distributions are not stable over time. As a
result, the future is unknown and unknowable. Thus, the Post Keynesian
concept of agency is one that must be consistent with the postulate of nonergodicity.
To speak of the Post Keynesian concept of agency is perhaps too generous, for this is an area of work that is still in development. Indeed, some
economists have advanced the claim that important aspects relating to the
concept of agency are completely lacking in Post Keynesian economics
(Hodgson 2001, p. 22). Still, a perusal of the work in this area clearly reveals
a set of factors that will undoubtedly be at the core of any Post Keynesian
concept of agency likely to develop in the near future.
Since agents make choices, they must possess a capacity that enables
them to accomplish this. The idea of making a choice involves more than
just a random or capricious action. To make a choice is to engage in an
intentional act based upon reasons and beliefs. Thus, agents must be
capable of having reasons and beliefs. Further, to act intentionally implies
that one is attempting to bring about some result. An attempt to bring
about a certain result is an attempt to structure the world in which the agent
lives. To do so, an agent requires the ability to conceptualize the world both
as it is, and as the agent would like it to be. Thus, agents must have the
capacity to formulate a conception of the world and a conception of what
a good life would entail.
Agents equipped with these endowments are then able to make choices.
What does the idea of making a choice imply? While not an uncontroversial question, the basic idea is that an agent in a given situation could have
selected an action dierent from that actually undertaken. This, however,
does not imply that the agent makes unconstrained choices. A key feature
of the Post Keynesian concept of agency is that agents make choices
within the context of a social structure, where by social structure we refer
to such things as rules, relationships and institutions. The introduction of
the idea of the social structure immediately raises a question: what is the
1
Agency
relationship between an agent and the social structure? Two traditional

answers have been given to this question. The first, the methodological
individualist position, advances the claim that structure is entirely the
result of individual actions. Thus, the structure is determined by individuals. The second, due to methodological collectivism (or holism), posits that
individuals and their actions are entirely determined by the social structure. A hallmark of the Post Keynesian concept of agency is a rejection of
both of these views. For Post Keynesians, individual agents are born into
a social structure that deeply influences, indeed partly constitutes, the very
nature of the agent. However, it is equally true that the actions of agents
help to reproduce and transform the social structure. Thus, agent and
structure are mutually dependent upon, but not reducible to, each other.
The fact that agent and structure are not reducible to each other means
that each possesses powers and capabilities that are not solely derived from
the other.
Agents, then, make choices in the context of a social structure. The fact
that the social structure partly constitutes the individual agent means that
the social structure does more than simply constrain the choices available
to an agent. Rather, the social structure partly determines who an agent is.
At a deep level, an agent is constituted by the meanings of the world he or
she both holds and transforms. For Post Keynesians, meaning is not an
objective fact about the world. Rather, meaning is both created and transformed as the result of social interaction within a social structure. Diering
social structures enable diering types of social interaction that engender
dierent meanings and understandings of the world, hence leading to
dierent individuals. Likewise, dierent individuals with dierent understandings of the world will help to bring about dierent transformations of
the social structure. Agents and structure are then engaged in a dynamic
process of reproducing and transforming each other. We thus see the
context for a frequently heard Post Keynesian expression, institutions and
history (time) matter. Moreover, the fact that the actions of agents will
bring about a transformation of the social structure also provides an explanation for the existence of non-ergodicity. This follows once we are able to
see that the future will be made by people on the basis of meanings that
they will freely create, though in the context of the social structure.
The idea of a dynamic interaction between agent and structure helps illuminate another important aspect of Post Keynesian economics. While the
world is non-ergodic, it nevertheless often remains fairly stable for significant periods of time. Keynes, and Post Keynesians, partly account for this
through the existence of conventions. In explaining how entrepreneurs
make investment decisions in a world where the future cannot be known,
Keynes expressed the view that agents tend to follow a convention by which
Agency 3
they project into the future the present state of things, unless there is some
specific reason for believing change likely. For Keynes, conventions are
essentially shared rules of behaviour that enable individuals to take actions
in situations where the future results of these actions are unknowable. From
a Post Keynesian perspective, conventions exist because they are one of the
elements agents use to give a coherent meaning to the world in which they
live. Thus, conventions actually help create the world, hence the future.
Moreover, conventions are formed on the basis of social interaction, which
helps us to understand what Keynes meant when he wrote in regard to how
expectations of the future are formed: We endeavour to fall back on the
judgment of the rest of the world which is perhaps better informed
(Keynes 1983, p. 114).
It is one thing to state that individuals and social structure partly constitute each other without being reducible to the other. It is quite another to
explain just how this can be. It is clear that we wish to avoid complete reduction of the individual to the social structure, and vice versa. To conflate the
individual with the social structure is to remove the possibility of free
choice. To reduce the social structure to the individual is to deny the independent existence of physical, chemical and biological forces. Neither of
these positions will do. However, to say that individuals and the social
structure only partly constitute each other implies that each of these possesses some capabilities that are independent of the other. What, if anything, explains these independent capabilities? Two very dierent positions
can be found in the literature.
The first is known as the Cartesian dualist position. According to this
view, it is simply in the nature of things that there exist both material and
intentional causes. The material causes deal with the physical, chemical and
biological. Intentional causes are the basis of human agency. The existence
of intentional causes, founded upon reasons, is what makes free choice possible. Intentional cause, itself, is a bedrock category in the sense that little
more can be said concerning what causes intentional cause. As the critics
of this position would state, intentional cause is an uncaused cause.
There are a number of diculties with this position. At the philosophical level, no generally accepted argument has been developed to explain
how these two dierent types of causes can interact and cohere with each
other. Perhaps more importantly, at least from the perspective of Post
Keynesian economics, is the fact that the Cartesian view is at variance with
the idea that individuals and the social structure partly constitute each
other. According to the Cartesian position, materialist and intentional
causes are independent, bedrock categories, with neither owing its existence
to any other factors. Under such a view, individual choices may be constrained by the materialist factors that explain the existence of the social
Agency
structure, but individuals and the choices they make are not (even partly)
constituted by the social structure. This is a view of the world that is more
akin to the neoclassical conception rather than the Post Keynesian conception of economics. Finally, the idea of an uncaused cause strikes critics of
this position as being unscientific in the sense that the idea rules out from
the outset any possibility of further investigation.
The second position found in the literature is newer, and we shall refer to
it as the evolutionary position (Bunge 1980). According to this view, materialist and intentional causes are not independent of each other. Rather,
intentional causes are emergent properties of the material world. The idea
here is that the human capability of intentional choice has evolved over
time through the development of materialist (physical, chemical and biological) forces. The term emergent is used in the following, somewhat
special, sense. While human intentionality has evolved from materialist
forces, it nevertheless possesses irreducible properties of its own. In other
words, while intentionality evolves from materialist forces, it cannot be
explained solely in terms of these forces.
While the evolutionary position avoids the strict dualism inherent in the
Cartesian approach, it too suers from a number of diculties. For
example, there does not yet exist an adequate explanation as to how intentions actually evolve from materialist forces. More problematic, from the
perspective of a social scientist, is the precise meaning of the term emergent. Several questions arise here. First, what exactly does it mean to state
that intention is only partly explainable in terms of materialist forces? Does
this imply that some aspects of intention are not explicable in terms of anything else? If so, does this not raise the same types of objections that were
raised against the Cartesian approach? If, on the other hand, intention can
be fully explained by materialist forces, then a serious question is posed as
to whether there really exists such a thing as free choice. While the writings
of those in the evolutionary camp are clearly sensitive to this issue, it is also
the case that the impression is often given that a complete explanation of
intention in terms of other causes is what these writers truly seek. How this
could be accomplished, in a manner that maintains the possibility of free
choice, is unclear. For example Hodgson, in criticizing the idea of an
uncaused cause, writes:
I also noted that chaos theory suggests that even if the world is deterministic, it
may appear as entirely spontaneous and free. On recent reflection, I now believe
that the admission of the possibility of an uncaused cause is not only unnecessary, for the reasons given in my Economics and Evolution book, but also untenable, for the reasons given here. The concept of emergence makes the
compatibility of determinism and free will possible, but that does not sustain the
notion of an uncaused cause. (Hodgson 2001, p. 39)
Austrian school of economics
While Hodgson here claims that free will and determinism are compatible, many fear that the true import of this claim lies in the analogy with
chaos theory. That is, what is really compatible is the appearance of free
choice and determinism, not the reality of free choice and determinism.
E J. MK
D C. Z
See also:
Babylonian Mode of Thought; Critical Realism; Expectations; Institutionalism; Nonergodicity; Time in Economic Theory; Uncertainty.
Bibliography
Bunge, Mario A. (1980), The MindBody Problem: A Psychobiological Approach, Oxford:
Pergamon.
Hodgson, Georey M. (2001), Structures and institutions: reflections on institutionalism,
structuration theory and critical realism (mimeo).
Keynes, John Maynard (1983), The Collected Writings of John Maynard Keynes. Volume XIV:
The General Theory and After, London and New York: Cambridge University Press for the
Royal Economic Society.
Lawson, Tony (1997), Economics and Reality, London and New York: Routledge.
McKenna, Edward J. and Diane C. Zannoni (199798), Post Keynesian economics and the
philosophy of individualism, Journal of Post Keynesian Economics, 20 (2), 23550.
Austrian School of Economics

The appearance in 1871 of Carl Mengers Principles of Economics (Grundstze der Volkswirthschaftslehre) marked the birth of the Austrian School
of Economics. After the publication of the book, Menger became
embroiled in the famous Methodenstreit with Gustav Schmoller, and thus,
unfortunately according to some, it was left to others to develop his
insights, most notably Eugen von Bhm-Bawerk and Friedrich von Wieser.
The development of the Austrian School was further carried on by Ludwig
von Mises and Friedrich von Hayek, and later by Ludwig Lachmann and
I.M. Kirzner. Other famous economists associated to various degrees with
the school include Fritz Machlup, Oskar Morgenstern, Joseph Schumpeter
and G.L.S. Shackle.
Mises and Hayek are possibly best known for their respective criticisms
of centrally-planned economies. In Hayeks case this took the form of
drawing attention to the dispersed, partial, continually changing, and frequently contradictory information possessed by dierent economic agents
in any advanced economy. Hayek argued that the nature of this information made it impossible for governments to direct economic activity with
any semblance of economic eciency. The problem was not merely a collection and computational problem, as information required interpreting
and was continually altering. Hayek argued that this problem also raised
questions concerning the possibilities for governmental intervention in capitalist economies, thus continuing what has become something of an
Austrian tradition of emphasizing the advantages of free markets.
However, as with all schools of economics, homogeneity cannot be
assumed. Vaughn (1994) claims that all contemporary Austrian economists
subscribe to two views. First, social phenomena are to be explained in terms
of the ideas and actions of individuals (methodological individualism).
Second, human action takes place in time and under conditions of uncertainty. This leads to an emphasis on individuals possessing dierent knowledge and expectations, and also on the importance of institutions for the
coordination of actions.
Given this, it would seem that the possibility of forging links between
Austrian economics and Post Keynesian economics appears very promising, given that the latter also emphasizes the importance of historical time,
uncertainty and expectations, and institutions. Further, of the various
founding schools of neoclassical economics, Austrian economics would initially appear to be the most congenial to Post Keynesians. Thus Menger
drew attention to the importance of time and error in economics (Menger
1976, pp. 67.) and, in an article on money, argued that the demand for
speculative balances formed a significant component of the demand for
money (Streissler 1973). Mises also emphasized the uncertainty of the
future (1966, p. 105), argued that the ideas of human action and time are
inseparable (pp. 99.), and dismissed the spurious idea of the supposed
neutrality of money (p. 398).
All this might suggest that any disagreements between the two schools
are merely cases of disagreements between friends. However, the history
of the relationship between the two schools is one of mutual indierence,
incomprehension, and even downright hostility. In part, this is no doubt
explicable in terms of historical precedent. Thus Keynes thought Hayeks
review of his Treatise on Money was carried out with insucient good
will, responding by characterizing Hayeks book Prices and Production as
an extraordinary example of how, starting with a mistake, a remorseless
logician can end up in Bedlam (1973, p. 262).
However, the dierences between the schools go deeper than lack of
good will and the questioning of sanity. In exploring these dierences, I
shall take Hayek as the spokesperson for the Austrian School, and
Davidson for the Post Keynesian School. The main issue I shall focus on is
the attempt, by each school, to incorporate uncertainty into their respective economic theories.
As Vaughn (1994) acknowledges, there are two strands in contemporary

Austrian economics. The first strand, represented primarily by Kirzner,
views Austrian economics as a necessary supplement to mainstream economics. On this account, Austrian economics is not directly concerned with
equilibrium states, but with the processes through which equilibrium may
be attained, with pride of place given to entrepreneurial discovery and creation. The second strand, represented by Lachmann (who greatly admired
the work of Shackle), concentrates on uncertainty and divergent expectations, and is highly suspicious of any reference to the idea of equilibrium.
In a discussion of Austrian economics, Davidson draws attention to the
tension between these two strands: Austrian subjectivists cannot have it
both ways they cannot argue for the importance of time, uncertainty, and
money, and simultaneously presume that plan or pattern coordination must
exist and is waiting to be discovered (Davidson 1989, p. 468). Davidson here
correctly recognizes a problem in attempting to fuse the two strands into an
Austrian view. If entrepreneurs are formulating plans under conditions of
uncertainty, with limited knowledge and divergent expectations, how can it
be assumed that somehow these numerous dierent plans become coordinated? In a recent attempt to resolve this dilemma, Kirzner argues that,
given uncertainty, entrepreneurs make mistakes. However, there are underlying objective realities [that] exercise their influence upon entrepreneurial
production decisions (Kirzner 1992, p. 34). Entrepreneurs may make mistakes, but it all comes out in the wash and thus entrepreneurs can still be
viewed as successfully steering the economy towards equilibrium.
It is dicult to recognize this as a satisfactory solution, if only because
it seems to fall foul of either of the two views that Vaughn claimed all contemporary Austrians subscribe to. Either some mechanism exists which
ensures coordination despite ignorance and uncertainty. In this case, rather
than explaining how coordination occurs through individual ideas and
actions, it is explained as occurring despite these ideas and actions.
Alternatively, coordination occurs through entrepreneurs coming to learn
what the underlying objective realities are, in which case there is no ignorance or uncertainty.
Turning to an Austrian perspective, Hayek, in his later works, levelled
three main charges against Keynes. First, he argued that Keyness General
Theory had been written in response to certain historical events, and was
thus not a general theory. Hayek argued that the mistaken view that the
theory was general led later economists to apply the theory in inappropriate conditions, leading to inflation. Hayek partially absolved Keynes
himself from this mistaken application, as he thought Keynes would have
been horrified by the results.
Second, Hayek objected to Keyness use of a macroeconomic form of
analysis, in particular to adoption of aggregate concepts. This objection

was in keeping with the Austrian preference for methodological individualism. Thus Hayek argued that the use of aggregate concepts such as the
level of inflation or the level of unemployment concealed the fact that
dierent markets could be performing dierently, and it was these dierences that mattered to individual decision makers. Individuals thus did not
think in terms of aggregate concepts, but in terms of the specific features
of the markets they were concerned with. It is worth noting that this objection would necessarily entail Hayek rejecting any rational choice model
assuming a representative individual.
Third, Hayek objected to what he termed Keyness rationalism. He thus
identified Keynes with the belief that human aairs and problems could be
fully known and resolved through the application of reason. I shall suggest
that Hayeks objection here still raises questions concerning the Post
Keynesian project.
In his initial discussion of Austrian economics, Davidson emphasized
that Post Keynesian economics assumes that economic decision makers are
operating in a non-ergodic environment. In fleshing this idea out Davidson notes that the Post Keynesian emphasis on uncertainty implies that
economic decision-makers recognize that todays probabilities (if any) will
not govern the future outcome (Davidson 1989, p. 479). In contrast, economists who believe the world is governed entirely by ergodic process
believe that the future is merely a statistical reflection of the past (p. 478).
In a subsequent article Davidson draws attention to the fact that reality is
transmutable, as todays human action can create a new and dierent
reality (Davidson 1993, p. 430).
The Post Keynesian emphasis on non-ergodicity, coupled with the transmutability of reality, entails uncertainty, where no relevant information
exists today that can be used as a basis for scientifically predicting future
events (Davidson 1993, p. 430). As entrepreneurs are acting under conditions of uncertainty, market coordination may not occur, and thus there is a
role for governments in creating the conditions whereby full employment
might occur (Davidson 1989, p. 474). Consequently, through institutional
and political changes, society can intelligently control and improve the performance of the economy compared with what would occur under laissezfaire(p. 430). This statement embodies a continued belief in the rationalism
that Hayek criticized in Keynes. An enlightened society, through the use of
reason and intelligence, can improve on the workings of the economy.
However, the problem with this is that the Post Keynesian emphasis on
uncertainty appears to leave society bereft of any policy guidelines through
which this can be achieved.
If the economy is not at full employment, there seems to be a major
problem for any government in trying to ascertain what policies might alleviate this situation. Given uncertainty, the government cannot rely on
attempting to implement policies that have worked in the past, as the past
is an unreliable guide to the future. Further, as there is no information available today that can allow predictions concerning the future to possess any
form of scientific credibility, then governments cannot be sure that any new
policy recommendations will not exacerbate, rather than resolve, the
problem. These diculties are compounded with the acknowledgement of
a transmutable reality. There will necessarily be a time dierence between
governments formulating policy, implementing policy, and any results of
this policy being achieved. Yet by the time the policy has been implemented
and the further time that results can be ascertained, the reality will have
changed from what it was (even if known) when policies were formulated.
Post Keynesian economists can, with considerable justification, criticize
the view in some Austrian circles that it is possible to emphasize both uncertainty and market coordination. However, it would also seem that the Post
Keynesian emphasis on uncertainty raises problems for the argument that
governments can resolve coordination problems. This, of course, connects
up with Hayeks other point, that Keynes did not advance a general
theory. Keynes may well have correctly identified problems of market
coordination when he wrote, and correctly identified policy instruments to
resolve them. However, given uncertainty, the past is a fickle guide to the
future and, given transmutation, the world is now a dierent place.
In conclusion, Post Keynesians have a valid point when they argue that
an emphasis on economic uncertainty raises problems for the assumption
that market coordination can occur in the absence of governmental intervention. However, it can also be argued that the emphasis on uncertainty
raises problems for the assumption that market coordination can occur
through government intervention. An uncertain resolution of the problems
raised by uncertainty, perhaps?
S D. P
See also:
Equilibrium and Non-equilibrium; Expectations; Non-ergodicity; Time in Economic Theory;
Uncertainty.
References
Davidson, P. (1989), The economics of ignorance or the ignorance of economics?, Critical
Review, 3 (34), 46787.
Davidson, P. (1993), Austrians and Post Keynesians on economic reality: a rejoinder to
critics, Critical Review, 7 (23), 371444.
Keynes, J.M. (1973), The Collected Writings of John Maynard Keynes. Volume XII: Economic
Articles and Correspondence, London: Macmillan.
10
Kirzner, I.M. (1992), The Meaning of Market Process: Essays in the Development of Modern
Austrian Economics, London: Routledge.
Menger, C. [1871] (1976) Principles of Economics, trans. J. Dingwall and B.F. Hoselitz, New
York: New York University Press.
Mises, L. von (1966), Human Action: A Treatise on Economics, 3rd edition, Chicago:
Contemporary Books.
Streissler, E.W. (1973) Mengers theory of money and uncertainty a modern interpretation,
in J.R. Hicks and W. Weber (eds), Carl Menger and the Austrian School of Economics,
Oxford: Clarendon Press, pp. 16489.
Vaughn, K.I. (1994), Austrian Economics in America: The Migration of a Tradition,
Cambridge: Cambridge University Press.
Babylonian Mode of Thought

The expression Babylonian mode of thought has been used in economics
and particularly in connection with Post Keynesian economics in an
attempt to identify a way of approaching economic analysis which is quite
dierent from the mainstream. We start by tracing the use made of the
term, and then discuss in more detail its meaning and significance.
But first we need to consider the term mode of thought. It refers to the
principles of knowledge construction and communication which underpin
choice of methodology, and indeed daily life: As we think, we live
(Whitehead 1938, p. 87). A mode of thought is the way in which arguments
(or theories) are constructed and presented, how we attempt to convince
others of the validity or truth of our arguments (Dow 1985, p. 11). It is
important to dig down to this level, beyond the methodological level, since
arguments about the relative merits of dierent methodologies (such as
Post Keynesian and orthodox) can founder through lack of recognition
that dierent modes of thought are also involved.
The term Babylonian was used by Keynes ([1933] 1972) in his biography of Isaac Newton, where he challenged the conventional understanding
of Newton as a rationalist, the first of the age of reason. Instead [h]e was
the last of the magicians, the last of the Babylonians and Sumerians, the
last great mind which looked out on the visible and intellectual world with
the same eyes as those who began to build our intellectual inheritance
rather less than 10,000 years ago (Keynes 1972, p. 364). Keynes contrasted
the way in which Newton applied introspection to his knowledge of the
history of scientific thought, as well as to experience, in order to arrive at
explanations for natural phenomena, on the one hand, with the rational
proofs he constructed after the fact, on the other.
The term Babylonian then apparently fell into misuse until introduced
to modern economics, as an approach to philosophy of science, in Stohss
(1983) note on the subject of Keynes on uncertainty. He argued that
Keyness ideas on uncertainty could be developed further on Babylonian
lines. He had picked up the Babylonian category from Wimsatts (1981) discussion in terms of the social sciences in general, in juxtaposition to
Cartesian/Euclidean thought. According to the Babylonian approach,
there is no single logical chain from axioms to theorems; but there are
several parallel, intertwined, and mutually reinforcing sets of chains, such
that no particular axiom is logically basic (Stohs 1983, p. 87).
Wimsatt in turn had developed the idea from Feynmans (1965)
11
12
Babylonian mode of thought
representation of what he called the Babylonian tradition in mathematics,

which involved a range of starting-points for arguments, and thus a multiple derivability of physical laws. Feynman contrasted this with the
Euclidean approach, which ties all arguments to a set of axioms, and argued
that the Babylonian approach was preferable for physics: The method of
always starting from the axioms is not very ecient in obtaining theorems
(Feynman 1965, p. 47). Indeed the context of this argument is a discussion
of the limitations of mathematics for physics: The mathematical rigour of
great precision is not very useful for physics (ibid., pp. 5670).
Following on from Stohs, Dow (1985; 1996) explored the nature and
implications of Babylonian thought in order to understand the dierent
underpinnings of mainstream economic methodology from those of the
methodologies of other schools of thought. Post Keynesianism being one
of those schools of thought, the idea of Babylonian thought came to be one
of the ways by which Post Keynesianism has become identified. Further, the
specification of Babylonian thought in relation to Cartesian/Euclidean
thought has been used in methodological discussion. This represented one
of a range of projects over the last decade or so to specify the philosophical and methodological underpinnings of Post Keynesian economics.
Feynman (1965) presented Babylonian mathematics as consisting of an
array of chains of reasoning, not tied to any one set of axioms, but governed
by the practicalities of the problem at hand. It is thus a realist approach to
knowledge. Since no one set of axioms can be relied on as being true, single
long chains of reasoning simply serve to compound any inadequacy in the
axioms. Rather than constructing a single general formal system, it is seen
as preferable to segment reality for the purposes of constructing a range of
partial analyses, which are incommensurate; if they were commensurate, the
arguments could be formally combined. One chain of reasoning might focus
on one segment of reality such that a particular variable is exogenous, which
is endogenous to another chain of reasoning. One chain of reasoning might
rely on statistical analysis, while another might rely on historical research,
for example. Euclidean mathematics, by contrast, is a closed logical system
built on one set of axioms using one, mathematical, method; it abstracts
from practical problems in order to generate universal solutions within the
domain of abstraction. The logical system is thus governed by internal rules
rather than reference to reality. This non-realist style of reasoning is also
associated with Descartes, hence the term Cartesian/Euclidean.
A Babylonian system of thought is a form of open system of thought,
rather than the closed system of Cartesian/Euclidean thought. In an open
system, the identity of all the relevant variables and relationships between
them is not known, and in any case the meaning of variables and their interrelations is subject to change. There is scope for creativity and discrete
13
shifts, as well as for stability. In a closed system by contrast all variables are
pre-specified, and categorized as endogenous or exogenous; what is not
known is assumed to be random. (A model within a closed system may be
open or closed; it is the knowledge of exogenous variables which closes the
system.) Extrinsic closure rules out anything but random disturbances from
outside, while intrinsic closure rules out any change in the variables within
the system or in their interrelations.
In order to satisfy these conditions for closure, Cartesian/Euclidean
thought is characterized by dualism and atomism. Duals are the allencompassing, mutually exclusive categories with fixed meaning typical of
closed systems. Variables are endogenous or exogenous; values are known
with certainty (or within a stochastic distribution whose moments are
known with certainty) or are not known at all; relationships are either
causal or random; economic agents are rational or irrational, and so on.
Atomism involves building up a theoretical system on the basis of the
smallest units, which are independent of one another and of the system of
which they are a part rational economic men.
Babylonian thought is neither dualistic nor atomistic. The categories used
to account for social life in an evolving environment are not seen as readily
falling into duals. Indeed vagueness of categories is seen to have the benefit
of adaptability within a changing environment where institutions, understanding and behaviour undergo change. In a system of thought with a
variety of incommensurate strands of argument, variables may be exogenous to one strand but endogenous to another. Knowledge is in general held
with uncertainty (by economic agents and by economists), so the analysis
points to degrees of uncertainty. Further, some strands of argument may
refer to individuals, and others to the group level, since causal forces may act
in either direction. Indeed individuals are not seen as independent, and their
behaviour may change as the environment changes. Institutions and conventions provide the stability to allow decisions to be taken in an uncertain environment. In other words the social structure is understood to be organic.
The concept of Babylonian thought accords well with ideas developed
later in the literature which have relevance for Post Keynesian economics.
The characterization of Babylonian thought outlined above follows from a
particular understanding of the nature of the real world as being organic,
that is, itself an open system. Babylonian thought is thus realist, and indeed
holds much in common with the critical realist approach to economics
(although not its philosophical foundations). While Lawson (1994) argues
that critical realism does not in itself provide the basis for identifying
schools of thought among those who adopt a critical realist approach, this
need not be the case (see Dow 1999). The Babylonian approach suggests a
basis for dierentiation in the form of realist ontology adopted whether
14
the economist understands the economic process in terms of production or

exchange, class or the rational individual, and so on. The case for the compatibility between Babylonian thought and critical realism was made by
Arestis et al. (1999) in response to Walters and Youngs (1997) critique; see
also Dow (1999).
Similarly, Babylonian thought provides a rationale for pluralism. It justifies both methodological pluralism (methodologists analysing a range of
methodologies) and pluralism of method (economists using a range of
methods). If the real world is understood as organic, not governed by universal laws, then there is scope for a range of methodologies. Further,
Babylonian thought specifically supports the use of a range of dierent
methods for dierent chains of reasoning. But, to be operational, both
forms of pluralism are moderated by the way in which the open system of
thought is specified. How the real world is understood will govern the particular choice of methodology, and in turn the range of methods to be used.
The original expression of the Babylonian mode of thought was misunderstood by some as the dual of Cartesian/Euclidean thought rather than generating a unified methodology, it was seen as encouraging methodological
diversity in the extreme sense of eclecticism. Cartesian/Euclidean thought
oers a closed axiomatic system, which yields certain conclusions given the
axioms. When Babylonian thought was understood as an open system
without axioms, with incommensurate methods and with uncertain conclusions, it was taken to imply the absence of methodological principles an
anything goes approach. It was associated with pure pluralism in the sense
of a range of methods with no appraisal criteria by which to assess them.
But this is a dualistic interpretation. By avoiding dualism, Babylonian
thought is not forced into an anything goes approach. Rather, some criteria are required by which to choose segmentations of the subject matter for
analysis, the chains of reasoning to pursue, and the methods employed to
pursue them. The subject matter is regarded as too complex to be fully captured in any one analytical system. So a range of choices as to methodology
is possible within a Babylonian approach. Since Post Keynesians have a distinctive ontology, a distinctive methodology follows, which diers from the
methodology of other schools of thought that also employ an open-systems
mode of thought. The corollary is that, while Post Keynesians can (and do)
argue for their own methodology and theories, they recognize that others,
with dierent ontologies, will choose dierent methodologies and theories.
While thought progresses within Post Keynesianism, there is also evidence
in the extent of Post Keynesian work which crosses boundaries with other
schools of thought, that an open thought system fosters creative synthetic
developments.
S C. D
Balance-of-payments-constrained economic growth 15

See also:
Critical Realism; Econometrics; Non-ergodicity; Uncertainty.
References
Arestis, P., S.P. Dunn and M. Sawyer (1999), Post Keynesian economics and its critics,
Journal of Post Keynesian Economics, 21 (4), 52749.
Dow, S.C. (1985), Macroeconomic Thought: A Methodological Approach, Oxford: Blackwell.
Reprinted in a revised and extended version as The Methodology of Macroeconomic
Thought, Cheltenham, UK and Brookfield, VT, USA: Elgar, 1996.
Dow, S.C. (1999), Post Keynesianism and critical realism: what is the connection?, Journal
of Post Keynesian Economics, 22 (1), 1533.
Feynman, R.P. (1965), The Character of Physical Law, Cambridge, MA: MIT Press. Page references are to the Penguin edition, 1992.
Keynes, J.M. (1933), Newton the man, in Keynes, Essays in Biography, reprinted in Keynes,
The Collected Writings of John Maynard Keynes, Vol. X, London: Macmillan for the Royal
Economic Society, 1972, pp. 36374.
Lawson, T. (1994), The nature of Post Keynesianism and its links to other traditions: a realist
perspective, Journal of Post Keynesian Economics, 16 (4), 50338.
Stohs, M. (1983), Uncertainty in Keynes General Theory: a rejoinder, History of Political
Economy, 15 (1), 8791.
Walters, B. and D. Young (1997), On the coherence of Post Keynesian economics, Scottish
Journal of Political Economy, 44 (3), 32949.
Whitehead, A.N. (1938), Modes of Thought, Cambridge: Cambridge University Press.
Wimsatt, W.C. (1981), Robustness, reliability and overdetermination, in M.B. Brewer and
B.E. Collins (eds), Scientific Inquiry and the Social Sciences, San Francisco: Jossey-Bass,
pp. 12463.
Balance-of-payments-constrained Economic Growth

The balance-of-payments-constrained growth model provides a Keynesian
demand-oriented explanation of why growth rates dier. This approach
stands in marked contrast to the neoclassical growth theory (whether of the
SolowSwan or the endogenous variety), with the latters emphasis on the
role of the supply side. The central tenet of the balance-of-paymentsconstrained growth model is that a country cannot run a balance-ofpayments deficit for any length of time that has to be financed by
short-term capital flows and which results in an increasing net foreign debtto-GDP ratio. If a country attempts to do this, the operation of the international financial markets will lead to increasing downward pressure on the
currency, with the danger of a collapse in the exchange rate and the risk of
a resulting depreciation/inflation spiral. There is also the possibility that the
countrys international credit rating will be downgraded. Consequently, in
the long run, the basic balance (current account plus long-term capital
flows) has to be in equilibrium. An implication of this approach is that
there is nothing that guarantees that this rate will be the one consistent with
the full employment of resources or the growth of the productive potential.
16
Balance-of-payments-constrained economic growth
The main elements of this approach are set out in Thirlwalls (1979)
seminal paper. The growth of exports is determined by the growth of world
income and the rate of change of relative prices. The growth of imports is
specified as a function of the growth of domestic income, together with the
rate of change of relative prices. Substituting these into the definitional
equation for the balance of payments, expressed in growth rate form, gives
the growth of domestic income as a function of the growth of world
income, the rate of change of relative prices, and the growth of net international capital flows.
If the impact of the last two on economic growth is quantitatively negligible (as empirically is the case), the growth rate of income consistent with
balance-of-payments equilibrium is given by yB z/x/, where , , z
and x are the world income elasticity of demand for exports, the domestic
income elasticity of demand for imports, the growth of world income and
the growth of exports. These two equations for yB are alternative specifications of what has come to be known as Thirlwalls law. It can be seen that
the key factor determining the growth of a country is the growth of the
exogenous component of demand, that is, exports, which in turn is determined by the growth of world markets. Thus, the model is an extension of
the export-led growth hypothesis, but where the balance-of-payments constraint is explicitly incorporated.
There are substantial dierences between countries in their values of
(and of ) and hence in how fast these economies can grow without
encountering balance-of-payments problems. The disparities in and are
interpreted as reflecting dierences in non-price competitiveness (for
example, dierences in the quality of goods and services, the eectiveness
of a countrys distribution network, delivery dates and so on). Thus the
supply side is important to the extent that these supply characteristics play
a crucial role in explaining the growth of exports and, hence, income. This
stands in marked contrast to the way in which the neoclassical approach
emphasizes the supply side, where technical change and the growth of the
labour input are the causal factors in the SolowSwan growth model and
the growth of capital (broadly defined) is the causal factor in the endogenous growth models.
A necessary condition for the balance-of-payments constraint to be
binding is that the rate of change of the exchange rate is ineective in determining the growth of exports and imports. If this were not the case, then
real exchange rate adjustments could ensure that the balance of payments
was brought into equilibrium at any given rate of the growth of income,
including the growth of productive potential. However, it should be emphasized that the balance-of-payments-constrained growth model does not
imply that changes in relative prices have no eect on the current account.

It may be that changes in these are sucient to bring a current account
deficit back into equilibrium when, for example, the economy is growing at
or near its balance-of-payments equilibrium rate, but they are unlikely to
be sucient to raise the balance-of-payments equilibrium growth rate, per
se. Given the multiplicative nature of the export and import demand functions, to achieve the latter would require a sustained real depreciation.
There are a number of reasons why this is implausible. First, there may
be real wage resistance, which makes it dicult for a continuous nominal
depreciation to be translated into a corresponding sustained real depreciation. Second, firms may price to market so that imports and exports are
unresponsive to any changes in the real exchange rate. Third, the values of
the price elasticities of demand may be so low that the MarshallLerner
condition is barely satisfied. If the absolute value of the price elasticities
sum to one, then the rule yB z/ holds, even if there is a substantial rate
of change of relative prices. Goods and services that enter into international trade are for the most part highly dierentiated and so their demand
curves are relatively inelastic. Firms compete for sales predominantly by
attempting to shift outwards the demand curve for their products through
increasing their non-price competitiveness, rather than by moving down
the demand curve through improving their price competitiveness. (See
McCombie and Thirlwall (1994, chapter 4) for a discussion of the empirical evidence.)
Thirlwalls law may be regarded as a dynamic version of Harrods (1933)
foreign trade multiplier. McCombie (1985) demonstrated that in a more
complex Keynesian model than Harrod used, Thirlwalls law could be more
generally regarded as the workings of the Hicks super-multiplier. An
increase in export growth from, for example, a position of current account
equilibrium would increase the growth of income directly through the
Harrod foreign trade multiplier. Moreover, at the same time, by generating
an increasing current account surplus, it allows a further increase in the
growth of other domestic components of demand to occur, thereby raising
the growth rate even further, until the basic balance is re-established. The
combined eect of these two mechanisms represents the operation of the
Hicks super-multiplier in dynamic form.
There have been an increasing number of studies that have tested this
approach to economic growth. The general methodology is to estimate the
value of and for a particular country from export and import demand
functions (which include relative price terms) using time-series data. In the
original studies, ordinary-least squares was used, but recently more sophisticated econometric techniques have been adopted, for example, those that
test for stationarity and cointegration of the data. From the estimates of
and , a value for the balance-of-payments equilibrium growth rate can be
18
Balance-of-payments-constrained economic growth
obtained using the expression for Thirlwalls law, yB z/. (Alternatively,

yB x/ is sometimes used.) The balance-of-payments equilibrium growth
rate, when calculated over a period of a decade or longer, is often found to
be very close to the actual growth rate and this has been confirmed by a
variety of statistical tests. It is also commonly found that the estimates of
the price elasticities in the export and import demand functions are either
small or statistically insignificant, or both. This provides further evidence
of the unimportance of price competition in international trade. See
McCombie and Thirlwall (1994) and the minisymposium in the 1997
edition of the Journal of Post Keynesian Economics (Davidson 1997).
Of course, not all countries will necessarily be simultaneously balanceof-payments constrained. At any one time, some countries (or trading
blocs) may be policy constrained, where demand management policies
have resulted in the actual growth of income being below the balance-ofpayments equilibrium growth rate. This occurred in the 1970s and 1980s in
some advanced countries where governments attempted to curtail the rate
of growth of inflation by using deflationary policies. Other countries may
be growing so fast that they are resource constrained, such as Japan in
the early postwar period. The problem is that the balance-of-paymentsconstrained countries find that their growth rates are eectively limited by
the growth of these policy- and resource-constrained countries. If, for
example, a particular country curtails its growth for policy reasons, its
major trading partners are going to find that their balance-of-payments
equilibrium growth rates fall. Their actual rate of growth will then be curtailed, regardless of whether or not the conditions in their domestic market
warrant this (McCombie and Thirlwall 1994, chapter 7).
The approach does not just apply to countries with national currencies,
but the principle holds also at the regional level (ibid., chapter 8). This suggests that the formation of a monetary union, such as the EMU (European
Monetary Union), will not remove the importance of export growth and
the balance of payments in determining the overall growth rate of a
country.
There have been a number of criticisms of this approach to economic
growth. McCombie and Thirlwall (ibid., chapter 5) contains a lively interchange with Peter McGregor and Kim Swales, which first appeared in
Applied Economics, over such issues as the direction of causation, whether
the model captures non-price competitiveness, and whether the law of one
price renders the model incoherent. (It is important not to confuse the small
variation in relative prices due to the reasons set out above with the neoclassical law of one price. The latter, with its assumptions of competitive
markets and that the price elasticity of demand of exports is infinite for
a small open economy, does imply that countries cannot be balance-of-

payments constrained. However, in practice, prices are determined in
oligopolistic markets and are sticky for the reasons noted above.)
Krugman (1989) rediscovered the law, which he termed the 45-degree
rule. This is because one countrys growth relative to all others will be equiproportional to the ratio of the income elasticity of demand for its exports
to the income elasticity of demand for its imports. The relationship between
a countrys growth rate and the values of and is interpreted in a neoclassical manner and not as reflecting the Harrod foreign trade multiplier.
Krugman develops a model based on monopolistic competition and
increasing returns to scale. The number of product varieties produced in a
country is assumed to be proportional to its eective labour force, where
the latter is taken to be a measure of resource availability. As a countrys
growth rate increases, so does the number of varieties it produces, and this
increases both its share in world markets and its value of . Hence, the latter
is assumed to be determined endogenously. If this were true, it would mean
that a faster growth of the UK would suddenly raise the growth of its
exports and reduce the income elasticity of demand for imports such as to
prevent a deficit from arising, with no downward pressure on the exchange
rate. This is implausible and contrary to the historical experience.
Crafts (1988) notes that if, for example, the UK had maintained its share
in its overseas markets, the hypothetical or constant-market-share income
elasticity of demand for its exports would have been comparable in size
with those of the other advanced countries (which all tend to be roughly
equal). Consequently, its hypothetical growth rate of exports would have
been the same as those of the other countries. Hence, using the constantmarket-share estimates of , it is argued that the UKs balance-of-payments
equilibrium growth rate is approximately the same as those of the other
advanced countries, including Japan. But all this shows is that if the UK
had matched, say, Japan in terms of its non-price competitiveness, its hypothetical balance-of-payments growth rate would have been the same as
Japans. But the fact is that it did not, and the estimates of the hypothetical
income elasticities have no relevance at all as to whether the UKs growth
was actually balance-of-payments constrained. All these critiques are
assessed in greater detail in McCombie and Thirlwall (1997).
In conclusion, Davidson (199091, p. 303) has summarized this approach as a significant contribution to Post Keynesian economic theory in
its demonstration that international payments imbalances can have severe
real growth consequences, i.e., money is not neutral in an open economy.
J.S.L. MC
20
Banking
See also:
Exchange Rates; Globalization; Growth Theory; International Economics; Multiplier.
References
Crafts, N. (1988), The assessment: British economic growth over the long run, Oxford Review
of Economic Policy, 4 (1), Spring, ixxi.
Davidson, P. (199091), A Post Keynesian positive contribution to theory, Journal of Post
Keynesian Economics, 13 (2), 298303.
Davidson, P. (ed.) (1997), Minisymposium on Thirlwalls law and economic growth in an open
economy context, Journal of Post Keynesian Economics, 19 (3), 31885.
Harrod, R.F. (1933), International Economics, Cambridge: Cambridge University Press.
Krugman, P. (1989), Dierences in income elasticities and trends in real exchange rates,
European Economic Review, 33 (5), 103146.
McCombie, J.S.L. (1985), Economic growth, the Harrod foreign trade multiplier and the
Hicks super-multiplier, Applied Economics, 17 (1), 5272.
McCombie, J.S.L. and A.P. Thirlwall (1994), Economic Growth and the Balance-of-Payments
Constraint, Basingstoke: Macmillan.
McCombie, J.S.L. and A.P. Thirlwall (1997), The dynamic Harrod foreign trade multiplier
and the demand-oriented approach to economic growth, International Review of Applied
Economics, 11 (1), 525.
Thirlwall, A.P. (1979), The balance of payments constraint as an explanation of international
growth rate dierences, Banca Nazionale del Lavoro Quarterly Review, 128 (791), 4553.
Banking
The behaviour of banks in the loan market is of major significance to Post
Keynesian analyses of a monetary production economy. When the money
supply is endogenous and the central bank sets the cost of wholesale funds
(the base rate), and accommodates bankers demands for liquidity at that
rate, the direction of causation between loans and deposits embodied in the
traditional exogenous-money/money-multiplier model is reversed. In Post
Keynesian models, loans cause deposits and hence bring money into existence as an integral aspect of the operation of the real economy. Thus
banks are able to advance the financing for investment without the necessity for saving to have been accumulated beforehand. It follows that investment expenditures can be constrained by credit rationing but not by a
shortage of saving. Banks and their ability and willingness to extend loans,
then, are a key determinant of increases in the level of employment and
output following an ex ante increase in the demand for investment goods.
To Post Keynesians, the banking system is in a constant state of innovating and evolving in response to the profit opportunities presented by the
interaction of economic conditions with regulatory constraints (Moore
1988, p. 31n; Minsky 1986, chapter 10; Chick 1992, chapter 12). The development of new financing instruments and techniques associated with liability management has been especially significant in the evolution of the
Banking
21
banking system. Liability management refers to the ability of banks to vary

interest rates to attract both wholesale and retail funds which can be used
to finance lending activity. This is in contrast to the asset-management
strategy of banks implied by the orthodox account, where banks must
either passively await new deposits made available by central bank purchases of government securities, or finance new loans with the proceeds of
the sale of other assets. Negotiable certificates of deposit, security repurchase agreements, retail cash management accounts and the interbank
market are examples of innovations which have increased the elasticity of
the supply of financing relative to regulations intended to limit such
responsiveness. However, although these innovations have increased the
ability of banks to provide credit on demand, the move to liability management has also increased their exposure to liquidity risk when they do so,
since balance sheets are expanded, with potentially volatile sources of
funds matching the issuing of new and relatively illiquid commercial loans.
Bankers decision-making processes around these liquidity issues are the
subject of some debate among Post Keynesian monetary theorists.
This debate can be framed as disagreement on the extent to which banks
are quantity takers in their loan markets. In the case of the horizontalist
position, as defined by Moore (1988), banks set the interest rate on loans as
a profit-maximizing mark-up over the cost of funds. The mark-up equates
the banks marginal cost of borrowed funds and marginal revenue of lent
funds. Bankers provide loans on demand at that rate: In their retail loan
and deposit markets banks act as price setters and quantity takers (Moore
1988, p. 55). A shortage of funds is met by borrowing in the wholesale
market, using the tools of liability management. As mentioned above, the
cost of funds in the short-term wholesale market is determined by the
central bank, which, as a market-maker, will typically fully accommodate
banks demands. Should the stance of monetary policy tighten, the central
banks supply of funds will be restricted and the base rate will rise. In this
scenario, the credit-money supply function is demand determined and perfectly elastic at the mark-up over the base rate, with the important implication that the liquidity preference theory of the interest rate is thereby
invalidated (ibid., pp. 197204).
Other Post Keynesians agree that the central bank implements monetary
policy through its control of the base rate, but disagree that banks are quantity takers to the extent required by the pure horizontalist position. They
argue that bankers respond to changes in their liquidity preference by
systematically price- and quantity-rationing credit. Thus, liquidity preference theory is essential to an explanation of how and why bankers vary the
price and availability of credit for any given base rate (see, for example, Dow
1996). In part, this dierence of opinion can be explained as somewhat
22
Banking
semantic by pointing out that the horizontalist position is overstated in

relation to the issue of the quantity rationing of credit. In fact, in the horizontalist account, banks impose credit limitations on all loan applicants
and meet unreservedly only the eective demand for loans. That is, before
banks supply loans on demand, applicants have already been subjected to
criteria which identify them as applicants to be either fully or partially
restricted (see Moore 1988, pp. 556). Fully restricted applicants are those
who fail to meet the minimum collateral and income and maximum risk
requirements set by the bank. Partially restricted applicants are those who
are approved for borrowing up to a specified limit. Thus, the supply of
credit by the banking system is perfectly elastic, up to borrowers allotted
credit ceilings (ibid., p. 337, emphasis added; see also Lavoie 1996). It is
therefore a matter of timing when does credit rationing take place?
which determines whether one argues that banks are or are not unqualified
quantity takers.
However, a perfectly elastic credit-money supply function rules out price
rationing at any particular base rate. This is justified by the need to clearly
distinguish the Post Keynesian exogenous interest rate analysis from the
loanable funds, endogenous interest rate, orthodoxy which predicts that an
increase in investment necessarily raises the rate of interest (Lavoie 1996,
pp. 2767). Lavoie (p. 279) argues that the base rate is indeed exogenous
with respect to the income-generating process, and that it is a separate issue
as to how bankers establish mark-ups over that rate. But he further insists
that Post Keynesians must argue that there is no compulsion for banks to
raise lending rates as economic activity expands. The opposing view is that,
at a given base rate, banks will supply more credit to individual borrowers
only if accompanied by an increasing loan rate, since borrowers debt/
equity ratios increase and hence lenders risk rises (Dow 1996, pp. 500503;
Minsky 1986, chapter 8). Yet it is certainly plausible that, in a period of
optimism and inflated expectations with respect to future stock or flow
returns from assets, bankers perceptions of this increasing risk may be
muted to the point of non-existence, at least within the relevant range.
Indeed, this point is made by Dow (1996, p. 501, Fig. 1) and similarly by
Minsky (1986, p. 193, Fig. 8.4). Moreover, the paradox of debt may apply
when considering not a single firm but industrial firms in aggregate.
Specifically, leverage ratios of firms may fall as profits rise with rising
investment during the upturn, frustrating firms plans to finance their
expansions with debt rather than equity, and eliminating the increase in
lenders risk (Lavoie 1996, pp. 2856). Furthermore, Lavoie (pp. 2924)
argues that liquidity preference goes by the name of animal spirits in the
horizontalist account, and so denies that there is a serious incompatibility
between the views of horizontalists and others. Nevertheless, Dow (1996,
Banking
23
pp. 5024) does not agree that the supply of loans is necessarily perfectly
elastic and suggests that, even if banks will supply loans to creditworthy
borrowers on demand for most of the business cycle, during a downturn
the rising liquidity preference of the providers of wholesale funds, as well
as the providers of the capital which banks must hold to meet risk-adjusted
capital requirements, will cause bankers liquidity preference, and hence
their mark-ups, to rise.
The use of liquidity preference theory is all the more imperative when the
business cycle is considered because it allows a cornerstone of Keyness
legacy the existence of uncertainty to play a central role in explaining
the periodic crises which beset capitalist economies. These crises are characterized by significant increases in loan defaults and interest rate margins,
and a collapse in the willingness of bankers to continue to extend credit,
entailing a strong desire to move to more liquid balance sheets (a rise in
banks liquidity preference). The upturn which eventually follows the crash
is characterized by the reverse of these events. Thus over the cycle, and independently of changes in the base rate, banks revise their views on creditworthiness and appropriate loan rates and, therefore, on quantity and price
rationing. Orthodox models of credit rationing due to market failure in the
form of asymmetric information between borrowers and lenders miss this
key point: there simply is no full-information case, so that default probabilities can only ever be subjectively determined. As Minsky (1986,
pp. 23940) stresses, the prudent banker is faced not by objective probability distributions of returns but by uncertainty and hence the necessity of
subjective evaluations of risks. These evaluations are not constrained by
depositor monitoring, due to the absorption of risk by the monetary
authorities (the classic moral hazard problem). Because the future is
unknown and uncertain, rather than probabilistically known and risky,
bankers and loan applicants may or may not suer from asymmetric information but they will typically be making decisions under asymmetric
expectations. That is, even if they have the same information (and even if
they have the same risk preferences), they evaluate that information dierently (Wolfson 1996, pp. 45051).
A bankers evaluation of information about particular borrowers, their
investment projects and their likelihood of repayment involves assessing a
number of factors in relation to the banks established standards or conventions. Factors such as the borrowing history of the applicant, the applicants debt/equity ratio, the value of collateral, and expected future cash
flows given the banks view of the macroeconomic environment during the
period of repayment, are all subjectively assessed. Whether or not these
assessments lead to the provision of funds, and at what price, is a function
both of conventions what has happened in the recent past and the extent
24
Bastard Keynesianism
to which it is expected that the past will be repeated and of the degree to
which a banker is confident in the assessment. When the bank perceives an
increase in the default risk of borrowers, due for example to a downturn in
the economy, or when its own liquidity preference rises, it raises both the
loan price and non-price requirements, so that some borrowers pay higher
rates at the same time as other borrowers those unable to meet the nonprice requirements are rationed (ibid., pp. 45260). Thus the loan supply
function is subject to shifts as bankers perceptions move between pessimism and optimism, with their liquidity preference correspondingly rising
and falling. Banks extensions of credit are virtually unconstrained in the
increasingly optimistic environment of the upturn, which leads them to
finance increasingly fragile debt positions (see Minsky 1986, chapter 9).
Central banks, as lenders of last resort, play an essential role in mitigating
the crises which result from the inevitable reversals of expectations, liquidity preference and cash flows relative to cash commitments which end the
growth of the debt pyramid. In short, banks are both destabilizing and
indispensable to a monetary-production economy: banking is a disruptive
force that tends to induce and amplify instability even as it is an essential
factor if investment and economic growth are to be financed (Minsky 1986,
p. 229).
G H
See also:
Central Banks; Credit Rationing; Endogenous Money; Finance Motive; Financial Instability
Hypothesis; Liquidity Preference; Monetary Policy; Money; Rate of Interest; Uncertainty.
References
Chick, V. (1992), On Money, Method and Keynes: Selected Essays, edited by P. Arestis and S.C.
Dow, New York: St. Martins Press.
Dow, S.C. (1996), Horizontalism: a critique, Cambridge Journal of Economics, 20, 497508.
Lavoie, M. (1996), Horizontalism, structuralism, liquidity preference and the principle of
increasing risk, Scottish Journal of Political Economy, 43 (3), 275300.
Minsky, H.P. (1986), Stabilizing an Unstable Economy, New Haven: Yale University Press.
Moore, B.J. (1988), Horizontalists and Verticalists: The Macroeconomics of Credit Money,
Wolfson, M. (1996), A Post Keynesian theory of credit rationing, Journal of Post Keynesian
Economics, 18 (3), 44370.
More has been written about Keyness General Theory than any other work
in economics in the twentieth century. It has a reputation of being a dicult
book to understand. Some of Keyness greatest supporters were initially
25
hostile in their reviews. The doyen of American economics, Paul Samuelson

found the book so confusing it took him 12 to 18 months to begin to understand it, and then only when it was put in mathematical form. Popularizers
had to simplify the book for mass consumption. In the process of simplification and interpretation, the General Theory was presented in a way that
was comfortable to those brought up on the microeconomic supply and
demand apparatus. Again there was a simple graphical presentation (either
ISLM or the Keynesian cross) and issues could be discussed in terms of
shifts of curves, and slopes and elasticities, and marginal changes of variables. Its pedagogical attractiveness and simplicity soon ruled the textbooks.
However, while economics acquired a new macroeconomic model that
was simple, easily grasped and teachable, it was also a model that omitted
many important aspects of Keyness ideas. Joan Robinson in 1962 called
this vulgarization of Keynes Bastard-Keynesianism (Robinson 1971,
p. 90). Sidney Weintraub called it Hicksian Keynesianism or Classical
Keynesianism. Hydraulic Keynesianism is another label used, as is NeoKeynesianism. The term neoclassical synthesis describes the process
through which Keyness General Theory was reconciled with pre-Keynesian
thinking. The years from the mid-1950s through to the 1960s were the
golden age of the neoclassical synthesis, which in substance was the linking
of the Keynesian incomeexpenditure system with neo-Walrasian general
equilibrium analysis. During this period, the terms neo-Keynesian and neoclassical seemed interchangeable. Yet in this context Keyness contribution
was relegated to imposing several price rigidities which, however useful for
policy purposes, were nevertheless theoretically trivial. It is this trivialization of Keynes that so oends his true disciples.
The process of simplification and systematization of Keynes began
almost immediately. One can see this clearly in the titles of the articles and
books published at the time. John Hickss key article was titled Mr. Keynes
and the Classics: A Suggested Interpretation. James Meades article was A
Simplified Model of Mr. Keynes System and Alvin Hansens book was A
Guide to Keynes (1953). Models consisting of a small number of simultaneous equations, presented by Hicks, Hansen, Meade, Brian Reddaway,
Roy Harrod, Franco Modigliani and Oscar Lange, came to symbolize the
Keynesian revolution. The model was popularly known as the ISLM
model. It seemed to summarize whatever substantive message the interpreters took from Keyness book in a system of simultaneous equations whose
properties were similar to the standard partial-equilibrium supplydemand
approach. Once the equations were translated into diagrammatic form (the
HicksHansen diagram) many of the apparent obscurities and ambiguities
seemed to vanish. Futhermore, it provided a logical basis for activist policy
proposals associated with the Keynesian revolution.
26
Paul Samuelsons response to the General Theory is instructive. While he

acknowledged that it was a work of genius, it was also (he claimed) obscure,
confusing, overly polemical and poorly organized. Keynes, he alleged, had
no genuine interest in economic theory, although by intuition he seems to
have stumbled on the right path. What was needed was to update and
systemize the framework through a system of mathematical equations.
Indeed, the ISLM model was needed to comprehend what the General
Theory was all about, something even Keynes may not have been sure of.
While the ISLM model was used for advanced students, a truncated
version called the Keynesian cross was used for introductory students.
Samuelson remarked that the intersection C(Y)I with the 45-degree
line gives us our simplest Keynesian-cross, which logically is exactly like
a Marshallian-cross of supply and demand (cited in Weintraub 1977,
p. 47).
The Keynesian cross and ISLM became the two most popular ways
through which students learned Keynesian economics. Wide dissemination of the Keynesian cross was achieved through the various editions of
Samuelsons introductory text. Relationships between aggregate variables
constituted the foundations of the macroeconometric models generated.
Aggregate flows of the economy were likely to grind out less than fullemployment output levels through various rigidities and imperfections.
Keynesians were not concerned with the structure of the economy, as all
that was needed was to change a few dials to maintain adequate levels of
aggregate demand. The most important dials were those associated with
fiscal policy; hence early Keynesians were often called fiscalists. Financial
markets and monetary policy were neglected.
As ISLM and the Keynesian cross became the dominant orthodoxy in
macroeconomics, other issues vanished from the mainstream literature. But
did these models convey the essential message that Keynes intended to
convey in the General Theory?
The dissenters from the mainstream interpretation of Keynes all agree
that ISLM and the Keynesian-cross models miss the essence of Keynes, but
provide contrasting views as to what that essential message is. Joan
Robinsons critique is primarily methodological. She denies the legitimacy
of using comparisons of equilibrium positions to analyse processes in actual
time, and contrasts models in logical time and those in historical time. The
Bastard Keynesians assume microeconomic foundations so that markets
behave as if they were Walrasian competitive ones, but agree that an
economy can come to rest at an underemployment equilibrium, or move to
full-employment equilibrium very slowly, due to deficient aggregate
demand. Hence they support demand management policies. Robinson
argues that one cannot fit Keynes into a neo-Walrasian framework because
27
it cannot handle historical time. Keynes did not think in terms of simultaneous determination: he was a cause and eect man; investment determined
saving, aggregate demand determined output and employment, and so on.
Robinson contrasts the Marshallian micro model of Keynes with the
Walrasian micro model of the Bastard Keynesians.
The General Theory stimulated the construction and testing of aggregative models. These models attempted to replicate the actual economys
behaviour through various systems of mathematical equations, the coecients of which were derived from historical data. Lawrence Klein played a
key role here in pioneering the path for a generation of quantitative
research in Keynesian macroeconomics. The economics profession devoted
substantial resources to the construction, estimation, testing and manipulation of these large-scale econometric models for forecasting and policy
analysis. All the large-scale US models had properties similar to the ISLM
model. While these models were called Keynesian the modellers seemed to
have been unaware of Keyness 1939 critique of Jan Tinbergens econometric methods and he presumably would not have favoured this development.
Textbook Keynesianism has been subject to frequent attack, with
perhaps Sidney Weintraubs (1977) critique being the clearest. Weintraub
sees Keynesianism as only tenuously connected to Keynes. The Keynesiancross incomeexpenditure analysis uses simple equations relating consumption to income, investment to the interest rate, and an equilibrium
balance equation. When solved it leads to an equilibrium income that may
or may not correspond to a full-employment level of income. However, in
this model, with its inflationary and deflationary gaps, the economy could
experience either inflation or unemployment but not both. It could not
accommodate stagflation. A Phillips curve was then grafted on to the
Keynesian-cross analysis and the Phillips curve became identified as an
important part of Keynesianism via Robert Solow and Paul Samuelson.
Yet one cannot find the Phillips curve in Keynes, and it would have been
inconsistent with his approach to find a sustainable empirical relationship.
Keynes would not have supported any stable, dependable, long-term relationship between inflation and unemployment. It was precisely these features of the model that were so successfully attacked by Milton Friedman
and the monetarists. Similarly it was the Keynesian macroeconometric
models that were attacked by Robert Lucas and the New Classical macro
economists. In both cases the attacks related to the Bastard Keynesians; not
Keynes. Leijonhufvud (1968) highlights this distinction clearly.
Weintraubs alternative Keynesianism, using aggregate demand and
aggregate supply analysis, would have avoided these confusions. Stagflation
is inconsistent with ISLM and the Keynesian cross, but not with aggregate
demand and aggregate supply. Stagflation is not inconsistent with Keyness
28
framework. Lorie Tarshis also provided an early text using this framework,
but the aggregate supply curve was developed in an imperfectly competitive
setting. This allowed an easier incorporation of the wage unit into the
analysis. Fiscalism is another concern for Weintraub. Sheila Dow notes that
Keynes for most of his life was an endogenous money person and a monetary theorist (Harcourt 2001, p. 48). The neglect of monetary factors in the
postwar period would not have been in the spirit of Keynes.
Paul Davidson (1972) and Hyman Minsky (1975) have both emphasized
the important role Keynes attributed to the financial sector. For Minsky,
Keynesian economics as the economics of disequilibrium is the economics
of permanent disequilibrium. He contends that the capital-asset valuation
process, in conditions of uncertainty, was central to Keyness argument and
that capitalism is inherently unstable due to its financial structure.
Davidson, along with George Shackle and others, focuses on fundamental
uncertainty. Decision making is undertaken in an uncertain environment,
and we need to examine the psychologies of the main players speculators,
investors, consumers, wage-earners. Keynes almost never refers to isolated
individuals; he speaks of the psychology of specific social groups. Davidson
tries to integrate the monetary detail of Keyness earlier A Treatise on
Money with the eective demand features of the General Theory. The overriding importance of uncertainty in investment and money markets means
that investment expectations are unquantifiable and unpredictable, and
their volatility influences the economys overall instability.
There are other critics of the Bastard Keynesians, like Victoria Chick,
who focus on the misapplication of the policy prescriptions of the General
Theory. Chick argues that fiscal stimulus was designed as shock treatment
and not intended to sustain an economy over a long time period. She maintains that Keynes would not have approved of fine-tuning and would have
supported the more selective use of fiscal policy, and not continuous budget
deficits. The excessive preoccupation with the short run has ignored the
long-run tendency to a lack of eective demand with involuntary unemployment. Less concern with fine-tuning and more attention to the longrun expansion of economic activity is required. Keyness mention of the
need for closer scrutiny of the level and composition of investment, or the
socialization of investment, has also gone unheeded by the mainstream
Keynesians.
To sum up the critique of the Bastard Keynesians, the so-called custodians of the real message of the General Theory claim that Keyness vision
is too rich to be encapsulated in one graph or a few equations. Wage stickiness or liquidity traps are not essential components of Keyness message.
The Economics of Keynes cannot be analysed in timeless, perfect information, general equilibrium models. A world of fundamental uncertainty
29
moving through historical time is essential to the message of Keynes.

Interpreting Keynes through ISLM is a distortion that forces the General
Theory into the older neoclassical mould. Mainstream Keynesians such as
Paul Samuelson and James Tobin reject these criticisms. Indeed, Robert
Solow approves of the Bastard Keynesian label because to him it suggests
hybrid vigour!
The debate between the Bastard Keynesians and the true disciples
has long been superseded by the attacks on Keynesians of any description
by the monetarists and, more powerfully, by the New Classical macro
economists. What one finds in the textbooks now is a far greater travesty
of Keynes than anything the Bastards did. For example, in Gregory
Mankiws Principles of Macroeconomics (1998) one finds a pre-Keynesian
loanable funds model; all a budget deficit achieves is higher interest rates,
crowding out and lower national saving. Mankiw is presenting the
Treasury View that Keynes so devastatingly attacked. In the unemployment chapter there is no mention of aggregate demand; the only causes of
unemployment are minimum wage laws, unions, eciency wages and job
search. Unemployment appears to be always at the natural rate, with deviations rare and transitory. A monetarist approach to inflation is provided.
This text concentrates on the classical principles of macroeconomics and is
a savage departure from the fundamental principles of any kind of
Keynesian thought. The Treasury View, the quantity theory of money
and Says Law are all found in modern guises. This is all presented as New
Keynesianism!
In this modern context, Georey Harcourt (2001) has made a plea for a
united front with Keynesians of whatever type, including the Bastard
Keynesians, to fight this external challenge. Despite fundamental dierences they have enough common cause to draw together to face the monetarist/New Classical assault. At times he says that this is a strategic alliance
and everyone is welcome at least at a tactical level, while at other times he
thinks that there is enough common ground for the dierences to be just
arguments within the family. His call for Keynesians of any persuasion to
coalesce into a united front seems to have mollified the earlier Post
Keynesian antagonism towards the Bastard Keynesians.
J L
See also:
Econometrics; Eective Demand; Keyness General Theory; Keyness Treatise on Money; New
Classical Economics; New Keynesian Economics; Stagflation; Time in Economic Theory;
Uncertainty; Walrasian Economics.
30
Bretton Woods
Bibliography
Davidson, Paul (1972), Money and the Real World, London: Macmillan.
Harcourt, G.C. (2001), 50 Years a Keynesian and Other Essays, Basingstoke and New York:
Palgrave.
Laidler, David (1999), Fabricating the Keynesian Revolution, Cambridge and New York:
Cambridge University Press
Leijonhufvud, Axel (1968), On Keynesian Economics and the Economics of Keynes: A Study in
Monetary Theory, London and New York: Oxford University Press.
Minsky, Hyman (1975), John Maynard Keynes, London and New York: Macmillan.
Robinson, Joan (1971), Economic Heresies, New York: Basic Books.
Weintraub, Sidney (1977), Hicksian Keynesianism: dominance and decline, in S. Weintraub,
Modern Economic Thought, Philadelphia: University of Pennsylvania Press, pp. 4566.
Young, Warren (1987), Interpreting Mr. Keynes: The ISLM Enigma, Cambridge: Polity Press.
Bretton Woods
The Bretton Woods agreement was part of the reorganization process following the Second World War. The Bretton Woods system, named after the
New Hampshire town where the conference was held in July 1944, was relatively short-lived. It extended from late 1946, when the declaration of par
values by 32 countries went into eect, to August 1971, if one takes the
closure of the gold window in the US as the end of the system, or late 1973,
if one takes the breakdown of the Smithsonian agreement. This was a
period of great prosperity. The Bretton Woods period exhibited the most
rapid growth of output of any monetary regime (Bordo 1993). As a result,
this period is usually referred to as the golden age of capitalism.
The Bretton Woods agreement is mostly known for imposing fixed, but
adjustable, exchange rates, on the basis of a golddollar system. Two main
characteristics of the Bretton Woods system should be emphasized, namely
the existence of a set of rules, that included fixed exchange rates, but also
capital controls and domestic macroeconomic policy autonomy, on the one
hand, and the hegemony of the US, on the other.
The Bretton Woods system represented the first successful systematic
attempt to produce a legal and institutional framework for the world economic system (James 1996, p. 27). It must be noted that the system worked
only because the US, the creditor country, was willing to pay the bill for
reconstruction through the Marshall Plan. The Marshall Plan, in turn, was
only possible in the environment of the Cold War. Post Keynesian authors
have emphasized the importance of creditor countries in expanding
demand on a global level (Davidson 1982).
The need for this legal framework was unanimously accepted as a way of
avoiding the negative consequences of the inter-war period financial speculation. Ragnar Nurkse forcefully presented the consensus view. According
to Nurkse (1944, p. 16), the flow of short term funds, especially in the thir-
Bretton Woods
31
ties, often became disequilibrating instead of equilibrating, or instead of

simply coming to a stop. This was partially true of the pre-1914 system, but
the main dierence was the absence of a hegemonic power capable of controlling capital flows through the variation of the interest rate (Kindleberger
1973). Further, according to Nurkse (1944, p. 22), in the thirties, there was
a gradual but persistent change in economic opinion. The price-level came
to be regarded more and more as a secondary criterion of economic stability. The state of employment and national income tended to become the
primary criterion. This change was to a great extent part of the eects of
the Keynesian revolution.
John Maynard Keynes was the chief British negotiator at the Bretton
Woods conference. It is important to note, given the prominence of
Keyness views during the conference, that he was not an advocate of either
fixed or floating exchange rates. His main objective was always the management of the exchanges to achieve domestic policy goals. The development
of the principle of eective demand led Keynes to support not only expansionary fiscal policies, but also low interest rate policies, whose ultimate
impact would be to reduce the significance of the rentiers income share, the
so-called euthanasia of the rentier.
To guarantee the euthanasia of the rentier, Keynes pointed out that the
central bank should be able to set the rate of interest independently from
any international pressures. Keynes especially insisted, during the long preparatory works and the negotiations for the Bretton Woods conference,
upon the idea that movements of capital could not be left unrestricted.
Keynes argued that we cannot hope to control rates of interest at home if
movements of capital moneys out of the country are unrestricted (Keynes
1980a, p. 276).
The fact that Keynes accepted, and even defended, the final agreement
at Bretton Woods, which diverged in several points from his Bancor proposal, can be attributed to the maintenance of capital controls in the final document (Crotty 1983). Also, it must be noted that, as much as Keyness Plan,
Harry Dexter Whites Plan also proposed the use of short-term capital controls (De Cecco 1979). The neoliberal thinking that had come to dominate
financial circles in the recent years had little standing at the Bretton Woods
conference. Keynes (1980b, p. 17) was categorical in saying not merely as
a feature of transition, but as a permanent arrangement, the plan accords
to every member government the explicit right to control all capital movements. What used to be a heresy is now endorsed as orthodox.
The control of capital flows means that the central bank does not need
to use the bank discount rate to attract inflows of capital, or avoid capital
flight. As a result the bank rate can be maintained as low as possible. A
reduction of the bank rate leads to a transfer from the finance or rentier
32
Bretton Woods
sector to the industrial capitalist and working classes, leading to an increase

in consumption and investment spending. In addition, low rates of interest
reduce the burden of debt servicing, so that active fiscal policies can be
pursued by the state without leading to an explosive increase of the debt to
GDP ratio. Thus, the prosperity of the golden age period is associated with
the euthanasia of the rentier, which was an integral part of the Bretton
Woods agreement.
Also, the inter-war period showed that greater capital mobility led to
greater exchange rate instability (Nurkse 1944). The fact that increased
capital mobility leads to higher volatility of exchange rates should not be
read as a one-way relationship. Flexible rates allow the opportunity for
speculators to profit from arbitrage; therefore profit-seeking speculation is
an inevitable outcome of the abandonment of fixed rates. In that sense, a
system of fixed but adjustable exchange rates is more conducive to a situation of reduced capital flows.
However, despite the intense preoccupation with capital controls, capital
movements began to play an important role in the late 1960s. A pool of
unregulated capital emerged as early as the late 1940s, when the Chinese
communist government placed its dollar earnings with a Soviet bank in
Paris. This was the origin of the so-called Euromarket (James 1996,
p. 179). However, it was not until the late 1950s, with return of the convertibility of the European currencies, and the removal of the current account
restrictions, that the transition from a dollar shortage to a dollar surplus
took place.
The growth of the Euromarket is also directly connected to the expansion of US multinational firms, and the consequent expansion of US
banking abroad. The collapse of Bretton Woods is related to increasing
speculative capital flows. According to Trin (1960) this resulted from the
fact that the US economy could not guarantee the convertibility of dollars
into gold at the fixed parity. In this view, the collapse of the Bretton Woods
system is directly connected to the increasing role of capital movements
and the incapacity of the hegemonic country to control them.
Whereas this view is incorporated in the conventional view, the main
cause of the demise of Bretton Woods is associated with the inflationary
pressures brought about by the expansionary fiscal policies in the US, and
the propagation of these inflationary pressures through the international
system. The increasingly expansionary fiscal policies of the 1960s resulting both from the Vietnam War and the Great Society experiment of the
KennedyJohnson administrations led to growing balance of payments
deficits. The US deficits were initially considered instrumental for the
working of an international monetary system that was desperately in need
of dollars to obtain the essential imports of capital goods needed for recon-
Bretton Woods
33
struction. However, by the late 1960s the accumulation of idle dollar balances started to put pressure on the money supply of the rest of world,
leading to inflation. That is, according to the neoclassical logic, inflation
was caused by the US fiscal and monetary policies, and transmitted to the
world as a result of the system of fixed parities.
The collapse of Bretton Woods, then, is related to the unwillingness of
foreign countries to import US inflation. That eventually broke the credibility of the fixed exchange rate commitments, and the willingness of the
several central banks to cooperate in order to maintain the fixed parities. In
other words, the Bretton Woods system failed because the fixed parity commitment was not credible in the face of accelerating inflation.
An alternative explanation for the inflationary pressures of the 1960s is
possible, though. This alternative explanation, compatible with the Post
Keynesian view, would minimize the eects of the US expansionary fiscal
policy in the demise of Bretton Woods. The golden age regime implied a
commitment to full employment and the creation of a safety net for unemployed workers. Additionally, the imposition of capital controls and the
cheap money policies which led to low real rates of interest implied a
favourable environment for workers. Parties with strong ties with the labour
movement were in power in several Western countries, and this was tolerated, to a great extent, since it was considered a form of reducing the
dangers of the Soviet menace. Furthermore, full employment tends to
increase the bargaining power of the working class.
In this environment, workers pressures for higher nominal wages would
usually be expected. For a given real rate of interest, and a fixed nominal
exchange rate, the only eect of rising wages would be higher prices. In
sum, inflation was the result of wage pressures (costpush) rather than the
expansionary fiscal and monetary policies (demandpull). In that sense, the
abandonment of the fixed parities is not connected to the loss of credibility in the face of higher inflation, since the causes of inflation lay somewhere else.
Post Keynesians emphasize the role of financial liberalization in the collapse of the Bretton Woods regime. Paul Davidson (1982) argues that the
US dollar represents the asset of ultimate redemption, and hence is used as
the measure of international liquidity. As a result the US benefits from a
more liberal financial system, since the centrality of US financial market
allows it to attract funds to finance persistent current account deficits. For
that reason, beginning in the 1960s the US adopted a more self-centred
financial policy, promoting financial openness in order to be able to face the
growing current account deficits. It is the increasing financial openness of
the 1960s, built into the American support for the Euromarket, that ultimately made the Bretton Woods system untenable. Whether the financial
34
Budget decits
disruption and the economic crises caused by the last wave of financial liberalization would lead to an eort to re-regulate financial markets is still an
open question. The echoes of Bretton Woods can be still heard in the calls
for a new financial architecture.
M V
See also:
Economic Policy; Exchange Rates; Globalization; International Economics; Stagflation.
References
Bordo, Michael D. (1993), The Bretton Woods international monetary system: a historical
overview, in M. Bordo and B. Eichengreen (eds), A Retrospective on the Bretton Woods
System, Chicago: University of Chicago Press, pp. 3108.
Crotty, James (1983), On Keynes and capital flight, Journal of Economic Literature, 21 (1),
5965.
Davidson, Paul (1982), International Money and the Real World, London: Macmillan.
De Cecco, Marcello (1979), Origins of the post-war payments system, Cambridge Journal of
Economics, 3 (1), 4961.
James, Harold (1996), Monetary Cooperation Since Bretton Woods, New York: Oxford
University Press.
Keynes, John Maynard (1980a), The Collected Writings of John Maynard Keynes. Volume
XXV: Activities 19401944, Cambridge: Macmillan and Cambridge University Press for the
Keynes, John Maynard (1980b), The Collected Writings of John Maynard Keynes. Volume
XXVI: Activities 19411946, London: Macmillan and Cambridge University Press for the
Kindleberger, Charles P. (1973), The World in Depression, 192939, Berkeley: University of
California Press.
Nurkse, Ragnar (1944), International Currency Experience: Lessons from the Interwar Period,
Princeton: League of Nations and Princeton University Press.
Trin, Robert (1960), Gold and the Dollar Crisis: The Future of Convertibility, New Haven:
Yale University Press.
Budget Deficits
Although budget deficits have usually been associated with Keyness economic policy proposals, he was not the first economist to put forward the
idea of utilizing deficit financing as a tool to fight unemployment.
Moreover, after he had written the General Theory, Keynes saw the deficit
only as an instrument of last resort. It was rather the Polish economist
Michal- Kalecki who persistently advocated the use of budget deficits.
Keyness early views on finance can be found in volumes V, VI and IX of
his Collected Writings, while his more mature outlook can be inferred from
volumes XXII and XXVII. Kaleckis writings on finance can be found in
volumes I, II and VII of his Collected Works.
Already in the 1920s Keynes realized that idle resources normally exist
Budget decits
35
in capitalist economies, and he favoured public works in order to increase

aggregate demand. He refined his viewpoint in the Treatise on Money,
where he argued that in an open economy with a fixed exchange rate regime,
variations in the interest rate could not induce simultaneously both internal and external equilibrium. Thus, with wage rigidity and a fixed exchange
rate regime, the unemployed would not tend spontaneously to be reabsorbed, while a decrease in the rate of interest to stimulate investment
would result in gold outflows and endanger the balance of payments. After
completion of the Treatise and while working on the General Theory he
came to utilize Kahns multiplier as a theoretical underpinning for his proposal, which was now openly related to budget deficits.
However, in the General Theory Keynes did not explicitly recommend
budget deficits, and in fact the government plays practically no role in that
book. The full incorporation of the government into Keyness mature
thought came later, while he was acting as an economic adviser to the
British government during the Second World War, and when he reflected
on what economic policy should be after the war was over.
Keynes believed that the fundamental role of the government was to
ensure conditions where uncertainty, which was the ultimate reason for
unemployment, would be minimized, so that enough private investment
would be forthcoming to absorb full-employment savings. Within this
context, the direct role of fiscal policy would be to smooth out undesired
variations in private expenditure (Carvalho 1997).
Keynes recommended separating the budget into two components, a
current and a capital budget. He considered unbalancing the current budget
as a last resort only, and argued that recourse to deficit would be a sign of a
failure of the governments overall economic stance. Moreover, in Keyness
view the capital budget should be balanced in the long-term but may be
adjusted to oset exogenous cyclical changes; the current budget may as a
last resort show surpluses or deficits to oset short-term failure of the
capital budget but also be balanced over the long term (Kregel 1985, p. 38).
Cyclical fluctuations could be prevented if two-thirds or three-quarters of
total investment were under public or semi-public auspices. Nevertheless,
Keynes opposed public works to stabilize the cycle because he considered
that public works had to be carefully planned. He also opposed the use of
taxation to aect consumption because he thought this method unreliable,
and he considered it politically infeasible continuously to alter the tax rate.
Michal- Kalecki, who independently discovered the theory of eective
demand, recognized very early the influence of government expenditure on
eective demand. In the early 1930s he showed that, below full employment, deficit financing would raise private profits, and would have a strong
expansionary eect. This can be demonstrated by assuming a closed
36
Budget decits
economy where all the productive sectors are vertically integrated and
where workers do not save. A rise in government expenditure financed by a
deficit will raise sales by an amount equal to the increase in government
expenditure plus the increase in workers consumption. Since the increase
in direct costs equals the increase in workers consumption, private profits,
which are equal to sales minus direct costs, will expand by an amount equal
to the rise in the budget deficit.
Somewhat later Kalecki added that financing expenditure by taxes levied
on private profits would also raise demand, since firms would not (or could
not) immediately curtail their investment expenditure after the tax rise had
been enacted, because investment was the result of previous investment
decisions which require a certain time to be completed and which would be
very costly to cancel, and because capitalist consumption is also rather
insensitive to expectations. Financing government expenditure with taxes
on profits would thus expand aggregate demand, provided that firms did
not pass on those higher taxes to consumers via higher prices. However,
since under this policy profits would not be raised, the expansionary eect
of government expenditure would be smaller than in the case of deficit
financing. Thus the notion that a balanced budget can be expansionary also
has its roots in Kalecki.
Deficit financing was one of Kaleckis Three ways to full employment
(1944), the title of a paper where he showed that, below full employment,
deficit financing would raise demand without necessarily entailing inflation, and that an appropriate monetary policy might also prevent the interest rate from rising with the higher output and employment consequent
upon the rise in the deficit.
Kalecki argued that in order to maintain full employment in capitalist
economies the deficit would probably have to grow continuously, but this
need not increase the burden of the national debt. The ratio of debt to
output need not rise, provided that the latter grew at a sucient rate (given
by the rate of growth of employment and labour productivity). Moreover,
for Kalecki even an increase in interest on the national debt as a percentage of output need not involve any disturbance in output and employment
if it were financed by a capital tax, levied on wealth (including ownership
of government securities). A similar result could be brought about with a
modified income tax, imposed on gross unearned income only (that is,
before deduction of depreciation) and where all investment in fixed capital
was deducted from the taxable amount.
Yet Kaleckis support for budget deficits was not unqualified. His misgivings, however, had nothing to do with the alleged evils the mainstream view
attributes to the deficit: inflation and crowding-out of private investment.
Rather, his reservations arose from his concern for social justice: the budget
Budget decits
37
deficit increases profits simultaneously with output and employment.

When income distribution is inequitable, and the political situation permits
full employment to be coupled with greater social justice, then expanding
government expenditure financed by higher taxes on profits and on income
of the higher-income brackets should be the preferred policy.
In the early 1930s, Kalecki was probably the first economist to show the
positive impact that deficit financing played in the German recovery under
the Nazi government. After 1945 he continued with the empirical application of his theory of public finance, and devised a novel and detailed methodology for analysing the eect of government spending on the level of
aggregate demand. This enabled him to demonstrate that the growth of
government expenditure financed by taxation on profits, not deficit financing, was at the heart of postwar economic recovery.
Deficit spending as a policy tool has recently been revived in the Post
Keynesian literature through the proposal to eliminate unemployment by
making the government the employer of last resort (ELR). This proposal
was originally put forward by Hyman Minsky (1986), who, however, tended
to oppose budget deficits. It has been developed more thoroughly by Wray
(1998), who recognizes that under an ELR strategy government expenditure will probably grow faster than tax receipts, and a budget deficit will
ensue. However, in his view the government can simply create enough new
money, or otherwise sell securities, to finance the deficit with an unchanging rate of interest. It has been shown that measures could be taken to minimize the tendencies to inflation and to external disequilibrium that might
arise if an ELR policy were implemented (Lpez 2000).
However, Steindl (1990) showed that in the most recent stage of capitalist development the budget deficit has turned from a policy tool to a consequence of the overall evolution of the economy. He utilized the
well-known identity:
(ISB)(XM)(GT)(SH H)
where (GT) is the budget deficit, (XM) is borrowing by the outside
world, (ISB) is borrowing by business, and (SH H) is lending by households. Steindl argued:
The budget deficit . . . used to be regarded as an active element, incurred on
purpose by the government. In the present circumstances it is more likely to play
a passive role, and be dominated by the other sectors. This is due to the large
share of taxation in an additional GDP, to the strong and quick reaction of consumers to a change in income and to the fact that the foreign balance is more
often dominated by outside influences than by domestic policy (by the GDP) . . .
The budget deficit is predominantly suered rather than contrived. (Steindl
1990, pp. 21718)
38
Business cycles
Nevertheless, as Godley persuasively argues, a budget deficit is still indispensable for long-run growth. He utilizes a single variable termed the fiscal
stance, which is given by dividing the total flow of government expenditure by the average tax rate. Godley prefers this variable to the budget
deficit, which measured ex-post facto is a bad measure of the impact of
fiscal policy because it notoriously fails to distinguish the eect of the
budget on the economy from the eect of the economy on the budget
(Godley and McCarthy 1998, p. 40). Utilizing the fiscal stance, Godley has
been able to closely track the evolution of the US economy in recent
decades.
J L G.
See also:
Economic Policy; Fiscal Policy; Kaleckian Economics; Keyness General Theory; Money;
Profits; Taxation.
References
Carvalho, F. (1997), Economic policies for monetary economies. Keyness economic policy
proposals for an unemployment-free economy, Revista de Economia Poltica, 17 (4), 3151.
Godley, W. and G. McCarthy (1998), Fiscal policy will matter, Challenge, 41 (1), 3854.
Kregel, J. (1985), Budget deficits, stabilization policy and liquidity preference: Keyness postwar policy proposals, in F. Vicarelli (ed.), Keyness Relevance Today, London: Macmillan,
pp. 2850.
Lpez, J. (2000), Budget deficit and full employment, Journal of Post Keynesian Economics,
22 (4), 54963.
Steindl, J. (1990), The control of the economy, in J. Steindl, Economic Papers, 19411988,
New York: St. Martins Press, pp. 21629.
Wray, L.R. (1998), Understanding Modern Money. The Key to Full Employment and Price
Stability, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Business Cycles
The time path of aggregate output and its main components exhibits significant fluctuations around trend values, as do other important variables,
including employment, productivity, prices, wages, interest rates and stock
prices. These fluctuations are recurrent but not regular. The pattern of comovements between the dierent variables, the amplitudes of the fluctuations and the length of the cycle vary over time. In fact, the delineation of
cycle from trend raises many problems, and cycles of dierent length may
coexist in the data; short-run fluctuations may take place with reference to
a long-run cycle, rather than around a constant exponential trend. The
term business cycles usually refers to relatively short cycles. This entry
therefore does not consider long waves. The possible interaction between
Business cycles 39
cyclical characteristics and any long-term trend is also ignored. The main
focus will be on fluctuations in output and employment for a closed
economy.
Business cycle theories can be categorized in dierent ways. One
common distinction concerns the exogeneity or endogeneity of the
cycles. In some theories the fluctuations are caused by external shocks and
the cycle, in this sense, is exogenous. The shocks may be completely
random and non-cyclical. They constitute the impulse, and the cyclical
pattern is produced by propagation mechanisms that spread out the eects
of the impulse. A positive shock, for instance, may induce firms to increase
their investment and, by raising the capital stock, this decision will aect
future conditions.
Michal- Kalecki (along with Ragnar Frisch and Eugene Slutsky) was a
pioneer of the external-shock approach. Most Post Keynesians, however,
have followed a dierent line. According to this alternative approach, external shocks may indeed hit the economy and aect movements in economic
activity, but fluctuations would occur even in the absence of shocks. The
fluctuations in this sense are created endogenously.
It is sometimes claimed that a reliance on external shocks leaves the cycle
unexplained and that endogenous theories are therefore intrinsically superior from a methodological perspective. The claim is not convincing.
Unforeseen shocks do hit the economy; some of these shocks must be considered exogenous, and it is easy to set up plausible propagation mechanisms that convert random shocks into irregular cyclical fluctuations.
It should be noted also that the distinction between exogenous and
endogenous cycles carries no implications for economic policy. Most neoclassical economists may take an external-shock approach, but policy intervention is both feasible and desirable in some models of exogenous cycles,
including some mainstream specifications. External shocks that require
policy intervention, moreover, also appear in Post Keynesian analysis, as in
the case of the desirability of compensating for autonomous shifts in
animal spirits. Conversely, endogenous cycles can be generated in models
in which markets clear and outcomes are Pareto optimal, as well as in Post
Keynesian models characterized by important market failures. Thus the
feasibility and desirability of policy intervention depend on the precise
structure of a theory and its cyclical mechanisms.
In general, Post Keynesian theories stress the instability of markets and
the need for both regulatory constraints and policy intervention. This
emphasis on the inherent problems and limitations of free markets, rather
than the exogenous/endogenous distinction, represents the substantive
dierence vis--vis most mainstream theories of the business cycle.
Endogenous business cycles can be generated in many ways, and at least
40
Business cycles
four distinct sets of mechanisms have been used in the Post Keynesian literature. The mechanisms are not mutually exclusive, and some contributions combine several mechanisms.
The determination of investment is central to theories that focus on the
goods market. As a main component of autonomous expenditure, high
investment leads to high levels of aggregate demand and output. A high
level of output, in turn, will be reflected in high rates of profitability and
capital utilization, and this will tend to induce high levels of investment and
output in the next period. If investment decisions are relatively insensitive
to changes in utilization and profitability, the resulting time path for output
will converge to a long-run equilibrium. A high sensitivity, on the other
hand, makes this long-run equilibrium (locally asymptotically) unstable:
following a slight displacement from the equilibrium position, the economy
does not return to the equilibrium but moves further away.
Local instability of this kind can be turned into perpetual fluctuations,
rather than cumulative and unbounded divergence, if there are appropriate
non-linearities in the investment function and/or in other equations of the
model. The existence of ceilings and floors represents a simple example
of such non-linearities (gross investment cannot be negative, for instance,
and output cannot exceed a full-employment ceiling) but other, less crude
non-linearities may also keep the movements bounded and convert local
instability into endogenous cyclical movements. Kaldor (1940) is a classic
reference for non-linear models in this multiplieraccelerator tradition, but
variations on this theme also characterize early contributions by Roy
Harrod, Michal- Kalecki, Paul Samuelson, John Hicks, Joan Robinson and
Richard Goodwin.
Investment needs to be financed, and financial markets are given a critical role in some aggregate-demand-based theories of the business cycle.
The financial instability hypothesis developed by Hyman Minsky (1982)
represents a prominent example. Suppose that, having recovered from
past turbulence, the economy now appears to be approaching a smooth
equilibrium path. Along this path expectations are largely being met and,
using Minskys terminology, there is financial tranquility: firms are able
to meet their financial commitments. This very state of tranquility will
induce changes in the risk assessments of both lenders and borrowers.
Risk premiums fall; lenders start giving loans they would previously have
rejected, and borrowers are increasingly prepared to finance their projects
in speculative and risky ways. These behavioural changes relax the financial constraints on the rate of investment and a boom ensues. Gradually,
the fragility of the financial system increases until a financial crisis
causes a rapid rise in interest rates and a contraction of credit and investment. A return to cautious financial practices now follows and the
Business cycles 41
process repeats itself (although the precise financial instruments and
institutions may be new and dierent). The result is perpetual, endogenous fluctuations.
The role of labour markets and income distribution has been emphasized
by a Marx-inspired literature, with Goodwins (1967) model of a growth
cycle as the most influential example. The model describes the dynamic
interaction between the distribution of income and the accumulation of
capital. When there is low unemployment when the reserve army of
labour is small, in Marxs terminology workers are in a strong position
and the real wage will be increasing. As real wages increase, however, profit
rates suer and the rate of accumulation declines. With a constant
capitaloutput ratio, the growth rates of output and employment fall, too.
Unemployment soon starts to increase, the balance of power starts shifting
against workers, and, when the balance has shifted suciently, the share of
wages stops increasing. Since the level of profitability is low, the rate of
accumulation will also be low and the rate of unemployment keeps rising
at this point. The capitalists now get the upper hand, the wage share starts
falling, and profitability and accumulation gradually increase. This increase
in accumulation gradually raises the rate of employment, workers once
again gain wage increases, and the cycle is complete.
This model formalizes Marxs general law of accumulation and, in
Goodwins original version, the model has no Keynesian features. It presumes that the capital stock is fully utilized at all times; output is determined by the supply side without reference to aggregate demand, and
investment adjusts passively to the level of saving. Hybrid models have tried
to overcome this weakness by including both Keynesian and Marxian features in the same model (Skott 1989).
Political intervention may itself be a source of fluctuations. This is a position stressed by many free market advocates, but the Post Keynesian argument for a political business cycle is dierent. The classic reference is a short
paper by Kalecki (1943). In a technical sense, Kalecki argued, governments
may have the ability to control aggregate demand at (near-) full employment,
but the maintenance of full employment generates cumulative changes in
worker militancy. Increased militancy and inflationary pressures quickly
bring together a powerful bloc of business leaders and rentiers and (supported by economists who declare that the situation is manifestly unsound)
the government allows unemployment to rise. The result, Kalecki argues, is
a political business cycle. Although applied by Kalecki to short cycles, the
argument is possibly better suited to deal with longer-term fluctuations, and
it has been used by a number of writers in relation to the rise in unemployment in the 1970s and 1980s.
Mathematical models have played an important role in the analysis of
42
Business cycles
business cycles in both Post Keynesian and mainstream theory. Not all Post
Keynesians are comfortable with the use of these formal techniques.
Business cycles, however, involve complex, dynamic interactions and in a
purely verbal analysis it is virtually impossible to keep track of these interactions and their implications. Without formalization it may be dicult to
decide, for instance, whether a given argument implies that there will be persistent fluctuations, explosive divergence or convergence to a smooth path.
Most formal models of endogenous fluctuations are deterministic. This
might seem a serious drawback. The empirical evidence shows irregular
cycles and, from a theoretical perspective, it should be easy for both private
agents and policy makers to forecast (and to take action to prevent) a cycle
that was regular and deterministic.
This objection is not as powerful as it might seem. First, the endogenous
view of cycles does not preclude external shocks, and the introduction of
shocks (ranging from natural disasters to policy shocks and changes in
animal spirits) may remove the regularity without aecting the underlying
cyclical mechanism. Second, deterministic, non-linear dynamic models can
produce chaotic outcomes that are hard to distinguish from those of a stochastic model. Prediction in these models is virtually impossible, since even
the smallest change in initial conditions has dramatic eects on the subsequent movements (for example, Day 1994). Third, the incentives for individuals to try to uncover and take into account aggregate regularities may
be small. Most decision makers face specific problems and uncertainties
whose eects on the outcome of their decisions dominate the eects of
movements in aggregate activity. With limited informational and cognitive
resources, these boundedly rational decision makers may choose to ignore
the possible influence of aggregate regularities altogether.
Post Keynesians, finally, have always emphasized the historical contingency of economic models. Structural and institutional changes, such as
the rise in the size of the public sector, the deregulation of the financial
markets, or increased international trade and capital mobility influence the
path of the economy and may necessitate a re-specification of the models.
The real-wage Phillips curve (a key element in the Goodwin model) may
shift, for instance, as a result of changes in labour market legislation.
Interestingly, the historical contingency is sometimes exaggerated in
popular writing; only a few years ago, for example, there was a widespread
belief that business cycles would disappear in the new economy. While
this belief has proved unfounded, the historical contingency and the complexity of business cycles have other implications. Small models, like the
ones described above, highlight particular mechanisms. But no single mechanism and no single source of shocks fully explain the diverse patterns of
fluctuations that have been observed, and the relative importance of the
Business cycles 43
dierent mechanisms may vary across both time and place. Thus the dierent models should be seen as useful tools rather than as complete explanations of the business cycle.
P S
See also:
Dynamics; Equilibrium and Non-equilibrium; Financial Instability Hypothesis; Investment;
Kaleckian Economics; Wages and Labour Markets.
References
Day, R. (1994), Complex Economic Dynamics, Vol. 1, Cambridge, MA: MIT Press.
Goodwin, R.M. (1967), A growth cycle, in C.H. Feinstein (ed.), Socialism, Capitalism and
Economic Growth, Cambridge: Cambridge University Press, pp. 548.
Kaldor, N. (1940), A model of the trade cycle, Economic Journal, 50 (197), 7892.
Kalecki, M. (1943), Political aspects of full employment, Political Quarterly, 14 (4), 32231.
Reprinted in M. Kalecki, Selected Essays on the Dynamics of the Capitalist Economy,
19331970, Cambridge: Cambridge University Press, 1971, pp. 13845.
Minsky, H. (1982), Can It Happen Again? Essays on Instability and Finance, Armonk, NY:
M.E. Sharpe.
Skott, P. (1989), Conflict and Eective Demand in Economic Growth, Cambridge: Cambridge
University Press.
Cambridge Economic Tradition

The starting-point must be Alfred Marshall (even though Maynard Keynes
called T.R. Malthus the first of the Cambridge economists and Keyness
successors were increasingly to draw on classical political economy and
Marx for inspiration). Marshall, though, was responsible for the foundation of the Economic Tripos (in 1903) and also, in large measure and at
least until very recently, for the approaches to economics in Cambridge
even as we know them today. The Marshallian tradition has it that economists should explain how the world works and then, if it does not work well
or fairly, do something about it (within well-defined limits). This should be
done by theorizing, doing applied work and formulating plausible policies.
The approach to applied economics emphasizes the importance of relevance in economics, incorporating the lessons of history, the institutional
context and previous social and political conditions, gathered under the
rubric of the rules of the game. Theory and measurement are interdependent, feeding back and modifying and expanding one another. This tradition has characterized the contributions of the Facultys Department of
Applied Economics, a research institute which started in 1945 with Richard
Stone (one of four Cambridge recipients of the Nobel Prize) as its first
director.
Marshalls major contribution was his huge Principles of Economics, first
published in 1890. It went through eight editions in his lifetime, as volume
I for the first five as he initially intended to write two or three more volumes.
What would have been the structure of Marshalls ideal Principles? In the
first volume he wrote about the nitty gritty of economic life what determines the prices and quantities of commodities produced, what determines
the wages, salaries and employment of dierent sorts of labour, what determines the rates of profit in various industries, that is, a theory of relative
prices and quantities. He introduced systematically into economics the use
of supply and demand functions and curves in order to analyse the formation of prices and quantities in, principally, freely competitive markets.
His second great contribution was to recognize in a deep way that time
is the most elusive, dicult yet relevant concept aecting economic life. To
try to capture this insight Marshall used three analytical concepts: the
market, the short and the long period. The first deals with existing stocks,
the last two with flows. The short period is an analytical device which takes
in a period long enough for employment and production but not for the
number of firms, or the amounts of machinery available and skilled labour
44
Cambridge economic tradition 45

to change; the long period is long enough for firms to enter or exit and for
the amount of machinery available and supplies of labour to change (the
methods of production known at the start of the long period, however, are
not allowed to change). These are not one-to-one descriptions of real life,
but analytical devices which exploit the concept of ceteris paribus. The
economist decides what may or may not vary, in order to get a grip on intricate interconnecting processes and so develop theories of prices and quantities of commodities, and of the services of the factors of production.
Money does not get a mention except as a ticket something in which to
measure things; it has little to no analytical role. Everything is done in real,
relative terms. Although Marshall understood general equilibrium analysis
and had a general equilibrium model in an appendix, he preferred to use
partial equilibrium analysis, examining one firm or one industry only, in
order to make the analysis manageable and obtain definite results (the limitations of which were explicitly stressed).
Money entered the scene properly when Marshall (in a never fully speltout second volume) developed the quantity theory of money in order to
describe what determined the general price level. He argued that, at least in
the long period, what was happening in the real sector and what was happening in the monetary sector of the economy banks and the financial
sector generally, the formation of the general price level were independent
of one another. Money was basically a veil. In the short period it was
admitted that monetary matters could have real eects, though this was not
worked out systematically because of the constraint of the dichotomy
between the real and the monetary. The role of monetary institutions,
including central banks, was to so control the monetary side of the
economy that the underlying real factors operating in a competitive environment were not handicapped in their determination of the allocation of
resources, with supplies and demands responding to each other and tending
to bring about a sort of social optimum.
This was only a sort of social optimum. The Marshallian tradition did
not contain an uncritical defence of laissez faire poverty, unemployment
and unsatisfactory working conditions were all recognized, along with a
limited role for government to tackle them. Nevertheless, logically, it was
required to argue that, if there was competition, there were strong forces to
ensure the production of goods and services that people wanted by businesspeople who were able to employ their capital as they wanted and
workers who could do the jobs they wished to. Only then was it possible to
argue that in the long period it was the quantity of money which determined the general price level, as long-period levels of activity and employment could now be regarded as givens along with the long-period value of
the velocity of circulation. As to the limitations of the outcome even in
46
Cambridge economic tradition
these circumstances it was A.C. Pigou, Marshalls successor, who developed the economics of welfare, analysing what happens, and what to do
about it, if social costs and benefits were not matched by their private
counterparts that the competitive system threw up. His influence is alive
and well today under the guise of externalities. Pigou, drawing on his
mentor, established another aspect of the Cambridge tradition, that our
subject should be, first and foremost, fruit-bearing rather than only lightbearing. This view is to be found in the many editions of The Economics of
Welfare. An interest in the causes of poverty and inequality, as well as in
the distribution of income, reflects this strand. It is especially associated
with the writings of James Meade, David Champernowne and Tony
Atkinson (who explicitly acknowledges Meades inspiration and example).
Keynes was Marshalls most distinguished pupil. He dominated Cambridge economics from the 1920s until his death in 1946, and beyond. His
work in the late 1920s and in the 1930s significantly extended and radically
changed the Marshallian tradition in which he was brought up. He was
driven, as were/are all the outstanding Cambridge economists, by an
intense seriousness: a desire to understand the world, especially why it malfunctioned, and how to make it a better place.
Trained as a mathematician, Keynes was also a fine philosopher as well
as a great economist. He always regarded economics as a branch of moral
philosophy. Three strands of his philosophical understanding are especially
relevant for the Cambridge economic tradition: first, that in a discipline
such as economics there is a spectrum of languages running from intuition
and poetry through lawyer-akin arguments to formal logic and mathematics, all of which are relevant for particular issues, or aspects of issues, in the
subject; second, that in the workings of complex economic systems, the
whole may be more than the sum of the parts; and third, two lessons
learned from Marshall: what are the principles which guide sensible (sometimes not so sensible) people doing the best they can in situations of inescapable uncertainty and what are the systemic eects of their behaviour?
The significance of these strands was made most explicit in the 1930s when
Keynes was writing The General Theory. In the 1920s and especially the
1930s he started to rethink drastically how the world worked, initially with
his close ally, Dennis Robertson, who is a bridge between Marshall and
Keynes, but who, in the end, tragically split with Keynes (a personal tragedy
for him and also a professional one for the development of economics).
As Keynes was rethinking Marshalls monetary theory, others at
Cambridge were starting to rethink (and in the case of Piero Sraa ultimately reject) Marshalls theory of the determination of prices and quantities at the level of the firm and industry. Sraa published two fundamental
papers in the mid-1920s, one in Italian and only recently available in an

English translation, the other in the Economic Journal (Sraa 1925 [1999];
Sraa 1926). Both contained an attack on Marshalls method, that is, the
extremely limited practical applications of partial equilibrium analysis
(and, he thought then but for dierent reasons, general equilibrium analysis). But he also suggested that monopoly rather than free competition was
the better model of how markets worked, that firms prices and outputs were
constrained by demand rather than by rising supply prices and costs. The
appropriate model was therefore one of mini-monopolies surrounded by
mini-monopolies so that they had to take account of their actions and other
firms reactions when setting prices. The 1926 paper helped precipitate the
imperfect competition revolution developed by Gerald Shove, Richard
Kahn and then Austin Robinson and, especially, Joan Robinson. Her 1933
The Economics of Imperfect Competition, though still Marshallian/Pigovian
in construction, greatly altered the emphasis and details of the results in this
tradition. Sraa had refuted the tradition by 1930 and had started on the
long trail which would lead through his edition (with the collaboration of
M.H. Dobb) of David Ricardos works (published between 1951 and 1963)
to Production of Commodities (Sraa 1960), both a critique of the foundations of neoclassical theory and simultaneously a rehabilitation of the
approach of classical theory including Marx. Joan Robinson only joined
him, more or less fully, in the postwar years, Kahn probably never fully and
Shove and Austin not at all.
Keynes became more and more dissatisfied with Marshalls way of
looking at the economy as a whole, especially the view that we could talk
about prices and quantities independently of what was happening in the
financial and monetary sectors generally. He also changed the emphasis
from the long period, the central core of Marshalls economics, to the short
period, including designing policies for other than that long run [in which]
we are all dead (Keynes 1923, p. 65). A Treatise on Money (1930), two
volumes, was meant to be Keyness magnum opus but it was too constrained
by the Marshallian tradition to be successful. So in the 1930s he started
again, aided by the remarkable group of young economists in the circus,
Kahn, James Meade, Austin and Joan Robinson, Sraa, as well as by Roy
Harrod in Oxford.
What did he do in his authentic magnum opus, The General Theory, published in 1936? The 1920s in the United Kingdom and then in the 1930s in
much of the advanced industrialized world was characterized by mass
unemployment. Economic theory, though, said that at least in the long term
it could not occur if impediments to competition were removed. Keynes,
working through his rational reconstruction of the traditional analysis,
decided that it was wrong, that there could be a failure of overall demand
so that people and machines could be involuntarily idle for considerable
48
periods of time and that there were not strong or indeed any forces at work
in an unregulated economy that tended to redress these situations. Why?
Principally because important expenditure decisions had to be made in situations of inescapable uncertainty about the future. This was especially true
of investment decisions, the desire to accumulate, which drove capitalist
systems along. Keynes showed that there were no persistent forces at work
which, at least on average, could produce enough investment to absorb the
resources released by what the community would voluntarily save at the fullemployment level of income, where all those willing to work under existing
conditions would have jobs. In situations of unemployment there was no
way in which those who were willing to work, but who were involuntarily
unemployed, could signal to employers that it would be profitable to employ
them. And indeed, it would not be profitable unless there were to be a simultaneous, autonomous rise in the total demand. It followed therefore that
there was a coherent logical case for government intervention; Keynes had
provided an explicit theory with which to rationalize the common-sense
policies which were being put forward at the time.
A barrier to this being perceived before was the real-monetary dichotomy, with money only a veil. But as it is also a store of value, people could
hold it and other financial assets rather than spend. In Keyness view this
second reason for holding money plays an important part in determining
the pattern of the rates of interest. The forces concerned may not be such
as to give a pattern which induces a rate of accumulation of real things
which osets full-employment saving. We now have an integrated theory of
the real and monetary, of a monetary production economy.
All members of the circus influenced Keynes but Kahn, his favourite
pupil, was especially influential in the making of The General Theory: first,
as a remorseless critic of the quantity theory as a causal explanation of the
general price level; second, through his work on the short period in the late
1920s, in which he made it a subject worthy of analysis in its own right
(though still at the level of the firm and industry); and, third, with James
Meade, through his 1931 article on the multiplier, which provided an essential concept for Keyness new system, showing how investment created
saving and not the other way round, as in the traditional view (Kahn 1931).
The General Theory (and a few following articles) were Keyness great
theoretical contribution to economics in the last century and the Cambridge contribution has built on these foundations ever since: developing
policies to run a war-time economy including keeping inflation in check,
using Keyness concept of the inflationary gap; designing the required international institutions for the postwar world at Bretton Woods, in order to
remove the contractionary, deflationary biases built into the operation of
much of the world economy (here Keynes leaves us, dead at the ridiculously

early age of 62 in 1946); and, third, in the postwar period developing longterm theories of distribution and growth over time. With this last it was
some of Keyness colleagues in the circus Kahn, Joan Robinson, Sraa
together with Nicholas Kaldor (who came to Cambridge from the London
School of Economics in the postwar years), Richard Goodwin and Luigi
Pasinetti, who were the pioneers.
Within the postwar development of growth theory their work stands out
as peculiarly Cambridge in that it draws on insights from the classical political economists, Marx and Keynes, initially in response to Harrods original and seminal writings just before and after the Second World War
(Harrod 1939). One basic question was whether capitalist economies could
maintain full employment of labour and capital over time when both the
employment-creating and the capacity-creating eects of accumulation
were taken into account, together with the classical concern with technical
progress, embodied through accumulation itself. In the Cambridge
approach (which includes Michal- Kaleckis contributions, principally
through Joan Robinson) investment led and saving responded, through
changes in both output and distribution, taking note of dierences in
saving propensities at the margin as between wages and profits (and their
recipients). (Kaldor flirted with being Jean Baptiste Kaldor for over a
decade (Samuelson 1964, p. 345), by assuming that growing economies
were fully employed and letting changes in distribution do all the work.)
The ultimate goal, probably only reached by Kalecki and Goodwin, was to
model descriptively the movement of industrial societies over time.
Joan Robinson and Kahn prefaced this objective with Golden Age analysis, walking before they ran, getting definitions and concepts clear and precise
before tackling the much harder task of disequilibrium dynamic analysis.
Ever impatient, Kaldors writings in the 1950s and 1960s were meant to be
descriptive analysis, theories to explain his famous stylized facts of economic growth, increasingly in the 1970s and 1980s by means of cumulative
causation processes. Neoclassical growth theories, though similarly stimulated (irritated?) by Harrods writings, tackled his conundrums the instability of the warranted rate of growth (gw), the unlikely correspondence of gw
with the natural rate (gn) by explicitly concentrating on the supply side and
the long-term eects of substitution possibilities in production. Initially
aggregate production function as well as multi-sector n commodity models
were used. This led to the Cambridge critique of capital and marginal productivity theory generally, associated especially with Kaldors, Joan
Robinsons and Sraas writings on value, distribution and capital theory.
Also, through Sraas contributions and Joan Robinsons writings on Marx
and her absorption of Kaleckis approach, the central classical/Marxist organizing concept of the surplus its creation, extraction, distribution and use
50
was integrated into the Cambridge approach and tradition. Pasinettis

(1981) theoretical essay on the dynamics of the wealth of nations is the most
systematic and comprehensive development of the classical and Keynesian
elements outlined above, thus making him the senior living heir of this strand
of the Cambridge tradition.
Finally, the tradition is marked by an interest in the history of our subject
and the relevance of our predecessors writings for current issues. Marshall
set the example; Keyness biographical essays reflect it (though he was not
always the most accurate or reliable historian of theory); and Joan
Robinsons writings are characterized by references to the insights of past
economists, often in order to back up her current interests and interpreted
accordingly! The two greats, though, are Kaleckis only two English gentlemen, one a communist and the other an Italian, Maurice Dobb and Piero
Sraa. Sraas edition of Ricardos works and correspondence (with the
collaboration of Dobb) and his attempt to rehabilitate the classical
approach in Production of Commodities are extraordinary examples of
scholarship and theory combined. Dobb was the foremost Marxist economist of his era and his writings and influence still diuse through modern
work, even when those aected are not aware of it. Dobb, together with
Phyllis Deane and Robin Matthews and their colleagues, also left a distinctive stamp on our understanding of economic history.
I have tried to make clear what I understand to be the Cambridge tradition. I have to say that many of the present decision makers in the Faculty
have done their best to suppress this tradition and to replace it with
approaches which reflect what they see as the best practice of leading
American departments.
G.C. H
See also:
Capital Theory; Joan Robinsons Economics; Kaldorian Economics; Keyness General
Theory; Keyness Treatise on Money; Sraan Economics; Treatise on Probability.
References
Harrod, R.F. (1939), An essay in dynamic theory, Economic Journal, 49 (193), 1433.
Kahn, R.F. (1931), The relation of home investment to unemployment, Economic Journal,
41 (162), 17398.
Keynes, J.M. (1923), A Tract on Monetary Reform, London: Macmillan (reprinted as volume IV
of Keyness Collected Writings, London: Macmillan for the Royal Economic Society, 1971).
Pasinetti, L.L. (1981), Structural Change and Economic Growth. A Theoretical Essay on the
Dynamics of the Wealth of Nations, Cambridge: Cambridge University Press.
Samuelson, P.A. (1964), A brief survey of Post-Keynesian developments [1963], in R.
Lekachman (ed.), Keynes General Theory: Reports of Three Decades, New York: St.
Martins Press, pp. 33147.
Sraa, P. (1926), The laws of returns under competitive conditions, Economic Journal, 36
(144), 53550.
Capital theory
51
Sraa, P. (1960), Production of Commodities by Means of Commodities. Prelude to a Critique

of Economic Theory, Cambridge: Cambridge University Press.
Sraa, P. (1925 [1999]), On the relations between cost and quantity produced, in L.L.
Pasinetti (ed.), Italian Economic Papers, Vol. 3, Oxford: Oxford University Press,
pp. 32363.
Capital Theory
Capital theory has been used as a shorthand term for the debate known
as the Cambridge capital controversies. The reference to Cambridge
follows from the dispute having mainly been conducted between prominent
figures, or figureheads, attached or aligned to Cambridge, Massachusetts,
especially MIT (Massachusetts Institute of Technology), and Cambridge,
England. The two Cambridges stood, respectively, as representatives for
mainstream neoclassical economics and its critics from radical political
economy. A key initiating text was penned by Joan Robinson (195354),
who was most closely identified with the early phase marking aggressive
popularizing of the critique. Sraas (1960) classic contribution, long
delayed in publication, has served as a basis both for a critique of neoclassical economics as was its intention and as an alternative to Marxist
value theory. The impact of the capital critique was at its height in the early
1970s, rising and falling with radical political economy. Today, as discussed
below, despite the Cambridge critique having won the debate and wrung
intellectual concessions from the mainstream, the latter proceeds in practice as if the controversy never occurred, replicating theoretical and empirical errors that were previously exposed and accepted as such.
The Cambridge critique raises a number of inter-connected issues. Here,
these will be reduced to, and represented by, three broad aspects. First, in
mainstream economics, it has been standard to represent an economy as
if it were reducible to an aggregate production function, F(K, L) say, where
K is capital and L is labour, with F() exhibiting the standard assumption of
decreasing but positive marginal products and overall constant returns to
scale. We place economy in inverted commas because it could stand for a
country, a sector, a single firm or any producing entity such as a household,
with a corresponding production function. In each case, even if imputed
for the household, it follows from a knowledge of F() and the capital and
labour in use, K and L, respectively, that the rate of profit, r, and the rate
of wages, w, can be determined by taking marginal products (with assumptions of full employment and perfect markets for inputs). In particular,
rf (k) where f is the per capita version of F; f (K/L)F(K, L)/L. It follows
that the rate of profit falls with an increase in capital per worker as a result
of the presumed diminishing marginal product of capital.
52
Capital theory
One way of interpreting the Cambridge controversy is in terms of

whether this stylized one-sector model (there is only one good, with the
capital input identical to the output) is capable of being representative of a
more complicated economy with more than one good. In other words, it is
a discussion about models specifically, are models with more than one
good reducible to an as if one-good model? The unambiguous answer is
no, unless special assumptions are made about the more complicated
economy that essentially make it equivalent to a one-good economy (all
outputs are produced with the same input proportions). The presence of
more than one good more or less completely undermines the results derived
from the one-sector model. First, distribution is not determined by technology alone what techniques are available and which are in use (by
analogy with the one sector, knowledge of F and the technique in use, that
is, which particular K/L). I have placed determined in inverted commas
because there is only a one-to-one association between technology and distribution if there is no double switching or reswitching, briefly elaborated
in the next paragraph. If you tell me which technology is use, I can tell you
distribution and, it should be added, vice versa. But this says nothing about
causation from one to the other or by other factors altogether.
Double or reswitching is dicult to explain briefly so an attempt will be
made to do so through use of present value curves that are assumed to be
familiar to the reader. Figure 1 shows a number of present value curves, T1,
PV
T1
T2
T3
r
0
Figure 1
A'
No reswitching
B'
Capital theory
53
T2 and T3, each representing a technique for producing output. Note, when
the rate of profit/interest (the two are treated interchangeably), r0, the
corresponding intercept on the present value axis is net output (gross
output over and above inputs without discounting). It might be presumed,
incorrectly, that higher net output (for T1 over T2 and for T2 over T3)
means higher capital intensity. But, even if this sort of one-dimensional
statement is to make sense (there are dierent capital outlays at dierent
points in time), it all depends upon the profile over time of streams of costs
and benefits. Even so, at low rates of interest, 0A, technique 1 would be
chosen, along AB, technique 2, and beyond B, technique 3. So a switch
is made from technique 1 to 2 at A, and from technique 2 to 3 at B. At any
rate of interest, there is a corresponding technique and vice versa. We could
also add more and more techniques, filling out an envelope, along which
there could be continuous switching from one technique to another as the
rate of interest changes.
PV
T3
C
T4
r
0
Figure 2
C'
D'
Reswitching
This is fine as far as it goes as long, as previously mentioned, as there is

no reswitching. The latter is illustrated in Figure 2. At C, technique 3
switches to technique 4 and, at D, technique 4 switches back to technique
3. There is no necessary one-to-one correspondence between technique in
use and the rate of interest. By knowing all techniques available and the
actual one in use, we are not able to determine the rate of interest/profit
54
Capital theory
as suggested by reference to the marginal product of capital for the as if

one-sector model (with inputs and outputs assessed at dierent points in
time making the present value curves equivalent to having more than one
good).
In addition, we cannot write down a sensible production function for the
economy as a whole since, even if we can do so for individual sectors in
terms of physical quantities of inputs and outputs, aggregating over sectors
will require an evaluation of the weight of each sector. This cannot be done
in a way that preserves the results of the one-sector model in particular,
that there is an inverse relationship between the measure of capital and the
rate of profit. With reswitching, as in Figure 2, it is possible for the same
technology to be in use at two dierent rates of profit. For the results of the
as if one-good model to carry over to the more complicated model, the
same physical quantities of capital would have to be measured as lower
(higher) for the higher (lower) rate of profit. In short, we cannot aggregate
capital sensibly to give rise to an as if aggregate production function.
The aspect of the capital controversy just covered is purely technical in
content the one-good model is not representative of a model with more
than one good. A second aspect concerns the empirical implications. For,
within mainstream neoclassical economics, it has been standard to estimate
the economy as if it were represented by a one-sector production function,
a practice often also to be found in use in radical political economy. This is
most notable in the residual method used to measure the contribution of
technical progress to economic growth, as opposed to the contribution
made by growth of inputs. In such work, even on the assumption of full
employment and perfect competition in all markets, capital is usually
aggregated by weighting it at current prices. As the previously reported
technical results suggest, this is entirely arbitrary. Indeed, because the
economy is being treated as if it only had one good, any change in prices,
for whatever reason and in favour of a capital good relative to a consumption good say, will be measured as if the quantity of capital had increased.
Accordingly, even though technology will not have changed, it will appear
as though output had remained the same despite use of more capital.
Technical progress will appear, falsely, to have been negative.
Again this is dicult to explain briefly. But refer to Figure 1 again. As we
go round the outside of the curves, the envelope traced by the available
techniques including others not shown, there is no technological advance,
just switching from one technique to another. But, associated with these,
for the as if one-good world, there will be changes in measured output
over and above (or below) those attached to technique. This is due to
changes in relative prices (which cannot be distinguished from changes in
quantities of input and output in the as if one-good world). In short,
Capital theory
55
changes in quantities and prices are treated as if they were purely changes
in quantities. It is simply empirical nonsense, as if the changes in the area
of a rectangle could be measured by reference to one side (as if it were a
circle with that radius, for example).
In short, the standard methods for measuring total factor productivity are
invalidated by the Cambridge critique. Further, this can be shown to have
nothing to do with reswitching the conflation of price and quantity eects
pertains even if no technique is preferred at two dierent rates of profit.
Despite this, empirical work on the basis of an aggregate production function has proceeded without regard to this totally destructive critique, with
only an occasional acknowledgement of the critique, often then with an
irrelevant and unexamined appeal to whether reswitching exists in practice.
The third aspect of the capital controversy concerns its wider significance for economic method and the choice between schools of thought. For
some, it seemed as if the whole of neoclassical mainstream economics were
invalidated by the critique. This is simply false, as the mainstream does not
depend upon a one-sector model of the economy. Indeed, the latter is a very
special case for, in the absence of more than one good, the model is eectively without demand (unless it be for choice of consumption over time)
and utility theory, a central component of the orthodoxy. The results of the
Cambridge critique can be accepted and incorporated within a model of
general equilibrium, in which there can be no presumption of simple
inverse relations between quantities of capital and rates of profit. For,
suppose there is an increase in physical quantities of all capital goods, the
rate of profit could rise if demand conditions are such that there is an even
greater increase in demand for capital-intensive goods or more capitalintensive techniques to produce them. What this does mean, though, is that
none of the intuitions attached to the one-sector model hold.
The debate was entirely conducted in terms of choice between linear
technologies. Essentially, the last aspect reveals that the debate is about
how to close such a model of supply. The mainstream can retreat into
general equilibrium and a utility-based demand theory. This, however,
leaves it floundering for a notion of capital other than as a physical quantity of inputs, initial endowments, that provides for a stream of utility. The
alternative oered by many from within the critique, as well as the critique
itself, has been inspired by Sraa. He showed that there is a trade-o
between the rate of profit and the level of wages, as previously posited by
David Ricardo. Hence, the terms Sraan and neo-Ricardianism are often
used interchangeably to suggest that technology (and technique in use)
alone do not determine distribution, only what is available to redistribute.
However, there is a dierence between these two closures in method. Sraa
(and Sraans) model in terms of physical conditions of production,
56
Capital theory
inputoutput matrices and the wageprofit trade-o they sustain

(although Sraas own preference for closure was via a profit rate related
to a distinct rate of interest determined by monetary factors). NeoRicardians retain an attachment to the labour theory of value. In either
case, with a linear technology, it is possible to close the system by appeal
to a trade-o between the rate of profit and the level of real wages, motivated by the idea that capital and labour confront each other over distribution. In addition, the Sraan approach has been used as a critique of
Marxist value theory, arguing that prices diverge from labour values, even
if modified by their transformation into prices of production.
As observed, capital theory attained its greatest prominence in the 1970s
when radical political economy was considerably stronger than today.
Despite the veracity of its empirical and theoretical results, and their acceptance by the mainstream (who conceded that empirical measurement of
performance ought to include both supply and demand and not just
supply), the mainstream now proceeds as if the Cambridge critique never
existed and shamelessly deploys aggregate production functions as if they
are without problems. This is particularly notable in the new or endogenous growth theory, where aggregate production functions are used for
theoretical and empirical work. It is a particularly appropriate symbol of
the analytical weaknesses of mainstream economics, and its ignorance even
of its own most recent history as a discipline, since the new growth theory
generally proceeds on the assumption that the economy as a whole can be
understood as if it were made up of a single sector (apart, occasionally,
from a separate sector to generate productivity increase in the as if onegood economy).
B F
See also:
Cambridge Economic Tradition; Growth and Income Distribution; Income Distribution;
Joan Robinsons Economics; Marginalism; Production; Sraan Economics.
Bibliography
Bliss, C. (1975), Capital Theory and the Distribution of Income, Amsterdam: North-Holland.
Fine, B. (1980), Economic Theory and Ideology, London: Edward Arnold.
Fine, B. (2000), Endogenous growth theory: a critical assessment, Cambridge Journal of
Economics, 24 (2), 24565, a shortened and amended version of identically titled School of
Oriental and African Studies (SOAS) Working Paper, No. 80, February 1998.
Harcourt, G. (1972), Some Cambridge Controversies in the Theory of Capital, Cambridge:
Cambridge University Press.
Harcourt, G. (1976), The Cambridge controversies: old ways and new horizons or dead
end?, Oxford Economic Papers, 28 (1), 2565.
Hodgson, G. (1997), The fate of the Cambridge capital controversy, in P. Arestis, G. Palma
and M. Sawyer (eds), Capital Controversy, Post-Keynesian Economics and the History of
Economics: Essays in Honour of Geo Harcourt, Vol. I, London: Routledge, pp. 95110.
Central banks
57
Robinson, J. (195354), The production function and the theory of capital, Review of
Economic Studies, 21 (1), 81106.
Sraa, P. (1960), Production of Commodities by Means of Commodities: Prelude to a Critique
of Economic Theory, Cambridge: Cambridge University Press.
Central Banks
Central banks are institutions with multiple functions. Existing central
banks are, in some cases, a spontaneous product of historical evolution, like
the Bank of England, or the result of social engineering, as in the case of
central banks created during the twentieth century, like the Federal Reserve
in the US. In both cases, central banks were created, or ended up with the
power, to manage the supply of national currencies. Orthodox economists
tend to see the main (or the sole) responsibility of a central bank as the
control of the available quantity of money in order to preserve its purchasing power. Keynesian economists, on the other hand, stress the fact that,
controlling the ultimate source of liquidity in a modern economy, a central
bank is responsible for the smooth operation of the financial system. For
Keynesians, thus, a central bank is first and foremost to operate as the
lender of last resort for the financial system, particularly for the banking
system, that creates means of payment under the form of demand deposits.
Post Keynesian economists, in particular, approach a modern market
economy as one organized around the existence of forward money contracts (see Davidson 1978). These contracts are essential to allow entrepreneurs to face the uncertainties that plague a market economy. Contracts,
however, create obligations to be discharged in the future by the delivery of
money from the debtor to the creditor. This simple fact is the foundation
of liquidity preference: facing the uncertainty of being able to pay debts
when they come due, the possession of money serves to lull ones disquietudes, to borrow Keyness expression. Money in these economies exists
either as currency or as its perfect substitute, demand deposits, private
liabilities that have the characteristic of being redeemable on demand, at
par. The acceptance of these substitutes, however, depends on confidence
that they will actually be redeemable on demand, at par. The only institution that can give this guarantee is the central bank, when it makes known
its willingness to trade bank deposits for currency. This makes the central
bank the lender of last resort to the banking system.
This view, proposed by Walter Bagehot in his famous Lombard Street,
was accepted by Keynes and is the foundation of the Post Keynesian
approach to central banking. There are some important dierences,
however, in the way the performance of this function is seen to constrain
central banks behaviour.
58
Central banks
Keynes himself believed that central banks, besides performing their

defensive function of lender of last resort, could also actively manage the
supply of money. In his Treatise on Money, Keynes stated that [t]he first
necessity of a central bank, charged with responsibility for the management of the monetary system as a whole, is to make sure that it has an
unchallengeable control over the total volume of bank money created by
its member banks (Keynes 1930 [1971], p. 201). He criticized the monetary
heretics who maintained that in some way the banks can furnish all the
real resources which manufacture and trade can reasonably require without
cost to anyone (p. 194). A bank performed a dual function, being a provider of money for its depositors, and also as a provider of resources for
[its] borrowing customers (p. 191). This meant that the creation of money
by banks had to take into consideration its eects on the value of money.
Central banks should use the interest rate as an instrument to control bank
reserves, since the aggregate volume of the deposits of the member banks
of a modern banking system depends on the reserve ratio which the
members aim at keeping, and the amount of reserves (in the shape of cash
and deposits at the central bank) (p. 43). Keynes recognized that the central
bank could be forced to supply reserves to banks, since there were situations
in which it had to purchase assets in virtue of an obligation, of law or
custom, to purchase such an asset if it is tendered on specified conditions
(p. 202). However, a central bank could compensate for these operations
and regain the initiative through the investments it could make: [t]he
amount of the central banks investments, since these are purchased and
sold on its own initiative, is entirely within its own control. Action directed
towards varying the amount of these is now usually called open-market
policy (p. 202).
In sum, for Keynes, a central bank should act as a lender of last resort
to the banking system, but this would not necessarily prevent it from also
controlling the volume of means of payment available in the economy. The
central bank uses the interest rate as an instrument to control the volume of
reserves of the banking system, but sets its aims both in terms of interest
rates and the volume of reserves.
Keyness approach was emphatically rejected by Kaldor. Kaldor argued
that Keynes never completely abandoned the quantity theory of money,
which led him to attribute a definiteness to the concept of money that it
lacked in modern economies and to give it an importance that it did not
have. According to Kaldor, money was not a precise concept, since the
public can use many instruments to make payments. Liquidity was a better
concept, but liquidity is an attribute shared by many types of vehicles in
dierent degrees. In Kaldors view, to single out money as a means of
payment would falsify the nature and operation of the payment systems,
Central banks
59
since it would imply confining the attribute of liquidity entirely to an arbitrarily chosen asset (Kaldor 1982). As a result, for Kaldor it was a mistake
to assume that a central bank could control the supply of money. It should
be concerned exclusively with interest rates, because these aect the actual
liquidity premium of the various assets by aecting the supply and demand
for each of them. Thus a central bank should set the interest rates over
which it had direct control, and freely supply the volume of reserves that
were demanded by banks at those rates. As a result, the money supply curve
should be represented by a horizontal line, in money/interest rate space,
originating at the level of the interest rate set by the central bank.
Kaldor argued that to perform the role of a lender of last resort did not
allow the central bank any room for pursuing an active reserve policy. Any
attempt to regulate the volume of reserves could threaten the solvency of
banks and thus put in jeopardy the supporting role of the central bank.
Kaldors later followers, like Basil Moore, extended Kaldors argument,
which was centred around the use of the discount window, to open-market
policies, suggesting, in opposition to Keynes, that central banks were constrained to supply a given amount of reserves also through open-market
operations. According to Moore, open-market operations are actually just
a way to make the discount window more ecacious: [o]pen-market operations are used not to aect the quantity of bank cash reserves . . . but
rather to compel banking institutions to make use of the central banks
accommodation facilities at the discount window, at the interest rates
charged by the central bank (Moore 1988, p. 89).
The horizontalist approach created by Kaldor represents, however, a
more extreme view that is not shared by all Post Keynesians. In fact,
authors like Paul Davidson, Hyman Minsky, Jan Kregel and Victoria
Chick advance approaches more faithful to Keyness own views, emphasizing, in dierent degrees, the capacity a central bank has of taking initiatives
in terms of reserve policies, while recognizing that one of its essential functions remains that of a lender of last resort. Davidson, for instance, defines
two strategies available for a central bank at any time in terms of creation
of reserves: (i) the income generating method, by which the central bank
accommodates the demand for reserves made by banks in order to satisfy
market demands for credit; and (ii) the portfolio change method, by which
the central bank takes the initiative to use interest rates to induce a desired
change in the amount of reserves at the disposal of banks (Davidson 1978).
In the first case, the central bank acts in a more passive way, validating the
demands coming from the credit market. In the second, though, the central
bank tries to implement its own strategies in terms of reserves, through
open-market operations, as Keynes suggested.
The debate between Keynesians and Kaldorians within Post Keynesian
60
Circuit theory
economics has been evolving for decades now, involving dierent conceptions of how a central bank should act, but ultimately relating to what each
strand believes the role of money in a modern market economy to be, how
the financial system works and evolves, and what is the role of monetary
policies, among other theoretical and empirical concerns.
F J. C C
See also:
Banking; Endogenous Money; Financial Instability Hypothesis; Kaldorian Economics;
Keyness Treatise on Money; Monetary Policy; Money; Rate of Interest.
References
Davidson, P. (1978), Money and the Real World, 2nd edition, London: Macmillan.
Kaldor, N. (1982), The Scourge of Monetarism, Oxford: Oxford University Press.
Keynes, J.M. (1930 [1971]), A Treatise on Money. The Applied Theory of Money, reprinted as
The Collected Writings of John Maynard Keynes, Vol. VI, London: St. Martins Press.
Moore, B. (1988), Horizontalists and Verticalists. The Macroeconomics of Credit Money,
Circuit Theory
The debate on Keynes has mainly focused on the principle of aggregate
demand and on the analysis of macroeconomic equilibrium with involuntary unemployment. This is in homage to the most widespread interpretation, which holds that Keyness innovative force exploded with the General
Theory (1936), the work in which he broke with neoclassical theory and
with most of his own earlier work. However, it is also possible to maintain
that the General Theory should be read as a continuation of the analysis put
forward by Keynes in A Treatise on Money (1930) and in other works before
and after the General Theory. According to this interpretation, Keyness
analysis should be considered part of the theory of the monetary circuit
(what Keynes called the monetary theory of production), which should
also include contributions from the first half of the twentieth century by,
among others, Knut Wicksell, Dennis Robertson and Joseph Schumpeter
(Realfonzo 1998).
In the second half of the century, starting in particular from the teachings of Keynes and Schumpeter, the theory of the monetary circuit was put
forward again and developed mainly by Italian- and French-speaking
scholars, such as Augusto Graziani, Marc Lavoie, Alain Parguez and
Bernard Schmitt. It has subsequently been supported by Franois Poulon,
Mario Seccareccia, Riccardo Bellofiore, Marcello Messori, Alvaro Cencini,
Claude Gnos, Elie Sadigh, Louis-Philippe Rochon, Giuseppe Fontana,
Circuit theory
61
Riccardo Realfonzo and others. The theory of the monetary circuit has
aroused growing interest, generating productive debate (for instance, see
Deleplace and Nell 1996) and further historical and analytical studies.
While significant dierences persist on specific points, most of the theoreticians of the monetary circuit follow substantially the same approach,
remaining well within the sphere of Post Keynesian theory.
As far as the basic analytical approach is concerned, the theoreticians of
the monetary circuit reject the methodological individualism typical of
neoclassical doctrine and adopt a socio-historical method. This does not
mean that their analyses are necessarily lacking in microfoundations: in
their view the study of individual behaviour is always subordinate to the
macro approach.
The simplest model of the monetary circuit, with a closed economy and
no state sector, can be described in the following way. Let us consider three
macro agents: banks, firms and workers. Banks have the task of financing
the production process through the creation of money, and of selecting
business plans; firms, through access to credit, buy factors of production
and direct the production process, making decisions on the quantity and
quality of output; workers supply labour services. The working of the
economy is described as a sequential process, characterized by successive
phases whose links form a circuit of money. A clear understanding of the
circuit theory can be obtained from Figure 3.
Banks
Initial finance
Final reimbursement
Firms
Purchasing of
labour services
Purchasing of
consumption goods
Savings
Workers
Figure 3
The monetary circuit
62
Circuit theory
The phases in the circuit are as follows:
1.
2.
3.
4.
5.
banks grant (totally or in part) the financing requested by firms, creating money (opening of the circuit);
once financing has been obtained, firms buy inputs; considering firms
in the aggregate, their expenditure coincides with the total wage bill; at
this point money passes from firms to workers;
once labour services have been purchased, firms carry out production;
in the simplest case, firms produce homogeneous goods;
at the end of the production process, firms put the goods on the
market. It can be envisaged that firms set the sale price following a
mark-up principle. Supposing workers have a propensity to consume
equal to one, firms recover the entire wage bill and maintain ownership
of a proportion (corresponding to the mark-up) of the goods produced. If the propensity to consume is less than one, once the workers
have purchased consumer goods they must make a further choice
about how to use their savings, either hoarding (increase in liquid
reserves) or investing (purchase of shares). If all the money savings are
invested in shares on the financial market, firms manage to recover the
whole wage bill;
once goods and shares have been sold, firms repay the banks (closure
of the circuit).
Starting from this synthetic description, the remarks below concern the
nature and role of monetary variables, the volume of production and
employment, the distribution of income and macroeconomic equilibrium
(Graziani 1989; Lavoie 1992; Parguez 1996; Realfonzo 1998).
According to the theory of the monetary circuit, money is a pure symbol
merely a bookkeeping entry (or a certificate) with no intrinsic value,
created by the bank in response to a promise of repayment. The bank is
defined as the agent that transforms non-monetary activity into activities
that are money. This approach therefore holds that it is the decision to grant
credit that generates deposits (loans make deposits). The money supply is
endogenous, in that it is essentially determined by the demand. On a theoretical level, the banking system could create money endlessly. In its turn,
the demand for money can be broken down into two distinct parts: the
demand for money to finance production (which Keynes called the finance
motive) and the demand for liquid reserves (dependent on the well-known
transactions, precautionary and speculative motives). According to the
theory of the monetary circuit, what mainly distinguishes entrepreneurs is
their access to bank credit. In fact, money as Schumpeter said is the
lever through which power over real resources is exercised. From what has
Circuit theory
63
been said, it follows that monetary circuit theory rejects traditional principles of the exogenous nature of the money supply and the neutrality of
money, as well as the quantity theory of money.
The volume of production is autonomously fixed by firms, based on the
expected level of aggregate demand. Naturally, production decisions taken
by firms may or may not be supported by banks. If there is credit rationing
by banks, firms are unable to translate their production plans into real production processes. To make the matter more complicated, it can be shown
that the production decisions taken by firms are also influenced by the possibility of equity rationing. One conclusion drawn by theoreticians of the
circuit is that the financial structure of firms is not neutral with respect to
production decisions. The employment level depends on firms production
decisions and therefore on the expected aggregate demand. The labour
market is thus described, according to Keyness teaching, as the place where
any shortage in aggregate demand is dumped (generating involuntary
unemployment). Macroeconomic equilibrium is compatible with the presence of involuntary unemployment.
According to the theory of the monetary circuit, as in Keyness original
work, in the labour market bargaining concerns only money wages. In fact,
the price level (and therefore the real wage) is known only at a later phase,
when workers spend their money wage in the goods market. This obviously
does not mean that, at the time when they bargain for their money wage,
workers have no expectations about the price level, but their expectations
are not necessarily confirmed by the market. Consequently, there may be a
dierence between the ex ante real wage (expected by workers) and the ex
post real wage (the actual real wage). If workers expectations about the
price level are confirmed, the expected real wage coincides with the actual
real wage. In a model with two types of goods (consumer goods and investment goods) this happens when voluntary saving equals investment. When
investment exceed voluntary savings there is a positive gap between the
expected real wage and the actual real wage which gives rise to forced saving.
As far as the firms profit is concerned, monetary circuit theory accepts
the thesis that firms as a whole earn as much as they spend. In fact, given
the mark-up (which in turn may be made to depend on the industrial concentration ratio), the higher the level of production (and therefore the
expenditure on inputs), the higher the firms real profit. In a model with two
types of goods consumer goods (wage-goods) and investment goods
decisions about the composition of production determine the distribution
of income. The higher the demand for and production of investment goods,
the higher the profits for firms. Thus scholars of the monetary circuit
reject the marginal theory of distribution in favour of a KaleckianPost
Keynesian approach.
64
Circuit theory
In the theory of the monetary circuit there is a strict distinction between

the money market and the financial market, and between the two interest
rates that are set there. In the money market, banks and firms negotiate and
the interest rate constitutes the price firms have to pay to obtain initial
finance. The money interest rate is basically a levy on the gross profit of
entrepreneurs. It should be underlined that, in the simple model here
described, firms can at the most repay the initial finance to banks but not
the interest as well. If there is hoarding, firms will not even be able to repay
the initial loan in money. It is possible to envisage that in this case firms may
decide either to settle their debt with the banks in goods, or remain indebted
to them. It is worth emphasizing, however, that this inability on the firms
part to repay the debt in money terms is not an inevitable feature of monetary circuit models. Indeed, as soon as one moves on to more complex
models with a state sector and/or an open economy, or to models in which
firms start production at dierent times (not simultaneously) this feature
disappears and in theory it is possible that, at the closure of the circuit,
firms are able to repay their entire debt (interest included).
In the financial market workers and firms negotiate and the interest rate
constitutes the price firms have to pay to raise the money not spent on the
goods market. It is, in fact, through the goods market and the financial
market working in conjunction that firms try to obtain the final finance, in
other words to recover the liquidity initially spent on purchasing inputs.
From what we have seen, it can be deduced that for circuit theorists there is
a sort of logical hierarchy between the money market and the financial
market. In fact, the financial market could not operate at all unless the
money market had already been operating. This means that while the individual firm can freely choose whether to get financing through the money
market or through the financial market, for firms as a whole no such choice
is possible. This is so unless there is a public spending deficit and/or a
surplus in the balance of payments such that enough money will flow,
through workers decisions to buy securities, into the financial market.
R R
See also:
Banking; Endogenous Money; Finance Motive; Keyness Treatise on Money; Money.
References
Deleplace, G. and E.J. Nell (eds) (1996), Money in Motion: The Post Keynesian and Circulation
Approaches, Basingstoke: Macmillan.
Graziani, A. (1989), The theory of the monetary circuit, Thames Papers in Political Economy,
Spring.
Lavoie, M. (1992), Foundations of Post-Keynesian Economic Analysis, Aldershot: Edward
Elgar.
Competition
65
Parguez, A. (1996), Financial markets, unemployment and inflation within a circuitist framework, conomies et Socits, 30 (23), 16392.
Realfonzo, R. (1998), Money and Banking. Theory and Debate (19001940), Cheltenham, UK
and Northampton, MA, USA: Edward Elgar.
Competition
Post Keynesians highlight the shortcomings of markets and the competition that regulates them. For them, as for Keynes, the competition of
markets does not make them self-adjusting. It does not keep the demand
for their products in line with the supply, or the supply in line with the
labour available for its production. Labour can be unemployed, and products in excess supply, under competitive conditions also, and while the competition of firms can bring down their prices, it cannot keep up their
production. Indeed, it may in fact ruin them.
The price competition of firms squeezes their profits, and this is the case
even if the wages of their workers fall with their prices, for their labour costs
are not the only costs of production. Firms have the expense of their plant
and equipment, and the service charges on their debt, and when those fixed
costs are high, and the investments of the firms irreversible, their price competition can wipe out their profit. It can drive prices down below costs,
expropriating the capital invested in their production (Eichner 1969). That
capital cannot be taken out of their industries when prices fall, and as long
as they stay above the average variable costs of production, the firms will
be better o selling products at a loss than not selling them at all.
While ruinous price competition is identified in economics with oligopoly, it is more in keeping with the conditions of perfect competition than
those of oligopoly. It is when the products of firms are homogeneous that
they must match the price cuts of competitors regardless of the costs, and
it is when their numbers are large, and market shares similar, that they
cannot eectively fix their prices. There is no dominant firm to enforce
their price-fixing agreements, and it makes little sense for the firms to
honour them when prices can be undercut and sales lost. And while perfectly competitive firms are small, and their productive capacity limited,
they can have fixed costs also. Indeed, their products could not be homogeneous if they were not produced by machinery, and their fixed costs will
lock them into a ruinous price war whenever their sales revenue falls short
of the amount needed for the recoupment of costs.
The perfect competition idealized in economics is far from ideal it
bankrupts firms and renders prices more volatile. The worst-performing
markets of the economy are the ones that come closest to the perfectly competitive markets of the economic texts (Kaldor 1985), and it is because
66
Competition
firms cannot operate under the price competition of those commodity

markets that they consolidate their industries and dierentiate their products (Eichner 1969).
The pricing power of firms is essential to their survival, and their investment depends on it also. They could not generate the revenue that investment requires if they were unable to hold up their prices, for profits would
not be high enough for the funding of investment (Eichner 1976). Firms
would not be able to finance investment with the profit from their products,
or secure (and repay) the loans needed for the external financing of investment projects. And since the profit from their products would be precarious as well as small, their long-run expectations could not be positive
(Shapiro 1998). They could not expect to make a profit on a product long
enough to recoup the costs of a plant and equipment or product development investment. Those long-lived investments are not profitable when
prices are variable and sales insecure, and, in the absence of the frictions
and restrictions of markets, there would be little investment in their products (Richardson 1990).
The entry barriers of industries increase and stabilize the investment in
products, and while they also lessen competition, they do not end it.
Oligopoly changes the competition of firms rather than eliminating it.
Their competition shifts from their prices to their products, and the product
competition of firms is as beneficial as the price competition. Indeed, it can
be more beneficial, for product competition increases the investment in
products.
The product changes that update and dierentiate the products of firms
require investment, as does the advertisement that publicizes the dierences
in their products. And the firms that compete on the basis of their products
must improve them as well as invest in them. They cannot increase market
shares without advances in the design or performance of products, and it
is only through those product innovations that they can maintain their
market dominance. The competition of new entrants is pre-empted not
through the prices firms charge for their products, but through the improvements made in them and their technologies.
The product competition of firms develops and improves products, and
the product competition of those that manufacture the equipment used in
industry improves production processes. Productivity improvements are
embodied in the products developed for the purposes of competitive
advantage. The productivity growth of modern industry is an outgrowth of
its competition, and it is because its firms compete on the basis of their
products rather than their prices that their competition advances the course
of technical progress. The oligopolization of industry endogenizes the
innovation process (Eichner 1976).
Competition
67
This is not to say that the oligopolization of industry is without cost. The
profit margins of firms can be too high as well as too low, exceeding the
level needed for the funding of their investment. Their investment is not
determined by their savings it also depends on their expectations and if
their profit margins rise without a corresponding rise in their investment,
the demand for their products will fall, reducing production and increasing
unemployment (Steindl 1976).
But the pricing power of the oligopolist is not absolute (Kalecki 1971).
It is limited by the competition of its industry, the substitutes for its products, and the risk of drawing new firms into its market. The substitution
and entry eects of its price increases limit the mark-up on its product, and
the firm will not increase prices in line with costs unless the prices of competitors also rise. There is no automatic transmission of costs into prices.
The fact that the price competition of oligopolists is tempered by the
conditions of their industries does not mean that there is no competitive
pressure on their prices. And while this competition may not be strong
enough to push prices down, and consumers cannot benefit from the productivity advances of firms unless prices stay in line with costs, mark-ups
can be maintained through wage increases as well as price reductions. If
money wages increase with the growth of productivity so will real wages,
and real wages will rise without the fall in prices that squeezes profits and
ruins firms.
The competition of firms does not have to be perfect for their prices to
be competitive and mark-ups constant. And while their competition is beneficial, there can be too much competition in industry as well as too little.
Perfect competition is not optimal.
N S
See also:
Equilibrium and Non-equilibrium; Innovation; Investment; Pricing and Prices.
References
Eichner, A. (1969), The Emergence of Oligopoly, Baltimore: Johns Hopkins Press.
Eichner, A. (1976), The Megacorp and Oligopoly, Armonk, NY: M.E. Sharpe.
Kaldor, N. (1985), Economics Without Equilibrium, Armonk, NY: M.E. Sharpe.
Kalecki, M. (1971), Selected Essays on the Dynamics of the Capitalist Economy, Cambridge:
Richardson, G.B. (1990), Information and Investment, Oxford: Oxford University Press.
Shapiro, N. (1998), Competition and employment, in P. Davidson and J. Kregel (eds), Full
Employment and Price Stability in a Global Economy, Cheltenham, UK and Northampton,
MA, USA: Edward Elgar, pp. 195201.
Steindl, J. (1976), Maturity and Stagnation in American Capitalism, New York: Monthly
Review Press.
68
Consumer theory
Consumer Theory
Few eorts have been made by Post Keynesians to explain how consumers
make choices. Does that mean that Post Keynesians accept the neoclassical
axioms of consumer choice? The answer is no. Although there have been
few contributions on consumer behaviour by Post Keynesian authors, there
is a certain degree of coherence among them. The few pieces that exist by
well-known Post Keynesians such as Joan Robinson, Luigi Pasinetti,
Edward Nell and Alfred Eichner fit, like a puzzle, with the rest of Post
Keynesian theory. These pieces must, however, also be tied to the work of
various institutionalists, social economists, marketing specialists and even
dissident mainstreamers.
The common ground of Post Keynesian consumer theory can be presented under the form of six principles (Lavoie 1994). They are:
1.
2.
3.
4.
5.
6.
the principle of
the principle of
the principle of
the principle of
the principle of
the principle of
procedural rationality;
satiable needs;
separability of needs;
subordination of needs;
the growth of needs;
non-independence.
The principle of procedural rationality asserts that agents lack perfect

knowledge and the ability to process a large amount of information. Agents
devise means to avoid complex calculations and considerations, and procedures enabling decisions to be taken despite incomplete information. These
means and procedures include rules of thumb, the acceptance of social conventions, and reliance on the hopefully better informed opinion of others.
Seen from the perspective of neoclassical substantive rationality, procedural rationality may seem to be ad hoc, but procedural responses are the only
sensible answer to an environment characterized by bounded knowledge
and computational capabilities, time constraints and fundamental uncertainty. It could also be called the principle of reasonable rationality.
In the case of consumer behaviour, it has long been established by marketing specialists that consumer choice usually involves very simple procedures (Earl 1986, p. 58). Very often there is no decision process to speak of:
purchases are made on recommendations, in conformity to social norms,
with the consideration of few alternatives, and on the basis of few criteria.
Some of the procedures that we follow are conscious we may then speak
of rules or conventions while others are unconscious; we may refer to
them as habits or routines, as in the case of a large part of our repetitive
spending on non-durable consumption goods.
Consumer theory 69
The second principle, that of satiable needs, can be likened to the neoclassical principle of diminishing marginal utility, but it takes a particular
meaning in the Post Keynesian theory of the consumer. Here satiation
arises with positive prices and finite income. There are threshold levels of
consumption beyond which a good, or its characteristics, brings no satisfaction to its consumer. Beyond the threshold, no more of the good will be
purchased, regardless of its price.
One has to carefully distinguish wants from needs, as do Lutz and Lux
(1979). There is a hierarchy of needs, where some are more basic than
others, which implies that they must be fulfilled in order of priority. In that
sense all needs are not equal. Some needs are bound to be satiated much
earlier than others. Needs are subject to a hierarchic classification and are
the motor of consumer behaviour. By contrast, wants evolve from needs.
They can be substituted for each other and constitute the various preferences within a common category or level of need (Lutz and Lux 1979,
p. 21). This leads to the next two principles of a Post Keynesian consumer
theory.
The principle of the separability of needs asserts that categories of needs
or of expenditures can be distinguished from each other. In the case discussed by Kelvin Lancaster (1991), with goods described by a matrix of
consumption technology with various characteristics, a separate need will
be associated with a submatrix of goods and characteristics arising out of
a decomposable matrix. The principle of the separability of needs is illustrated by the widely-used econometric models of consumer demand, which
assume that broad categories of expenditures enter separately into the
overall utility function. In the utility-tree approach, the principle of separability is pushed one step further, since these broad categories of expenditures are further subdivided into several branches.
The separability of needs allows the consumer to divide the decisionmaking process into a series of smaller multi-stage decisions. The consumer first makes an allocation of his or her budget among needs, and then
spends that allocation among the various wants or subgroups of each need,
independently of what happens to the other needs. Changes in the relative
prices of goods within a given category of wants will have no eect on the
budget allocation among various needs, while a fall in the overall price of
a group of goods corresponding to a given need will have repercussions on
the budget allocation of all needs. The principle of the separability of needs
imposes substantial restrictions on the neoclassical principle of price substitution, since separability severely limits the degree of substitutability
between goods in dierent groups. Indeed, a substantial amount of empirical evidence shows that general categories of consumption expenditures
have quite negligible own-price elasticities and cross-elasticities.
70
Consumer theory
Further restraints may be added if one goes beyond the principle of separability of needs, by introducing a fourth principle, the principle of the
subordination of needs. With this principle, utility cannot be represented by
a unique catch-all utility measure; it can only be represented by a vector.
The principle of the subordination of needs is often associated with the
notion of a pyramid of needs a hierarchy of needs as described by the
humanistic school of psychology (Lutz and Lux 1979). The integration of
the principles of separability and subordination leads to Nicholas
Georgescu-Roegens principle of irreducibility. Needs are irreducible.
In the case of utility-tree analysis, the first-stage budgeting problem is
resolved by assuming that money is allocated first to necessities and then to
discretionary needs. There is no substitution between the budget categories
apportioned to necessary needs and discretionary ones. All the principles
previously invoked culminate in this hierarchy: needs are separable and the
most basic needs are first taken care of in their order of priority, until they
are satiated at some threshold level.
There have been some formal representations of the above principles.
Hierarchical behaviour is known under the name of lexicographic preference
ordering, owing to its similarity with searching for a word in a dictionary.
Strict lexicographic ordering, however, is unlikely, and more sophisticated
lexicographic approaches have been suggested, with consumers setting
targets and thresholds, that is, with the addition of the first principle of
Post Keynesian consumer theory, that of satiation (Earl 1986). These noncompensatory ordering schemes are not only reasonable but also compatible
with procedural rationality, since a complete utility map is not required.
Decisions about the most basic needs can be taken quite independently of
the informational requirements of the higher needs. Consumers need know
nothing whatsoever about the prices of the goods that are part of the higher
needs, and they need not rank alternatives which they cannot attain or which
are beyond their satiation levels (Drakopoulos 1994).
Neoclassical authors deny that needs are subject to the principle of subordination. This, it must be presumed, is mainly due to the devastating consequences of the irreducibility of needs for neoclassical theory and its
substitution principle. Irreducible needs imply that they are incommensurable and therefore that everything does not have a price. A trade-o is not
always possible. The axiom of Archimedes, so popular with choice theorists, does not hold any more (Earl 1986, p. 249), nor does the axiom of
gross substitution (Eichner 1987, p. 632), so often invoked among general
equilibrium theorists. This reinforces the arguments of Paul Davidson
against the use of such an axiom in macroeconomics.
Having assumed that indeed there exists a hierarchy of needs, how do
consumers move up the steps of the pyramid? The basic answer is that indi-
Consumer theory 71
viduals move upwards in the hierarchy due to income eects. Beyond the
principle of satiation, lies the principle of the growth of needs our fifth
principle.
When a need has been fulfilled, or more precisely when a threshold level
for that need has been attained, individuals start attending to the needs
which are situated on a higher plane. There are always new needs to be fulfilled. If they do not yet exist, consumers will create them through innovation, but this may take time (Gualerzi 1998). Needs, however, often require
income to be satisfied. To go from one level of need to another dictates an
increase in the real income level of the individual. The fulfilment of new
needs, and therefore the purchase of new goods or new services, is thus
related to income eects. This is the microeconomic counterpart of the
Post Keynesian focus on eective demand, that is, on macroeconomic
income eects. What is being asserted is that income eects are much more
important in explaining the evolution of expenditure on goods than are
substitution eects. The latter play only a minor role in a static analysis of
consumer behaviour, when similar goods or goods fulfilling the same wants
are being considered. Indeed, changes in relative prices have an impact on
budget allocation between needs only in so far as they have an impact on
real income.
The sixth and last principle is the principle of non-independence. The
emphasis of traditional theory on substitution eects also has led to the
neglect of the learning process in consumption theory. How do consumers
rank their new spending opportunities? How do they learn to spend their
additional spending power? Consumers watch and copy other consumers.
Preferences are not innate, they are acquired by experience and by imitation of the consumption pattern of friends or of people of higher ranks in
the consumers hierarchy. Fads leading to large sales of specific products
are thus explained by the informational content of consumption by neighbours, relatives, friends or acquaintances. The impact of socio-economic
contact on purchases reinforces the belief that the composition of demand
depends on socio-economic classes. Decisions and preferences are not
made independently of those of other agents. A households pattern of
consumption will reflect the lifestyle of the other households that constitute its social reference group. Marketing ocers, through publicity, will
attempt to make sure that households follow the appropriate lifestyle.
Three of the main consequences of this Post Keynesian analysis of consumer choice may now be noted. First, macroeconomic models based on
the analysis of income classes and on income eects are a legitimate outgrowth of a Post Keynesian theory of the consumer where price substitution eects are not important or severely constrained to goods which
respond to similar characteristics, and where increases or changes in
72
Consumption
demand are mostly determined by increases in real incomes or changes in

consumer preferences. Second, the Post Keynesian theory of consumption,
based on the hierarchical nature of needs, is also reminiscent of the classical distinction between necessaries and luxury goods and of the Sraan
distinction between basic and non-basic commodities. Under these circumstances, to ignore substitution eects, based on relative prices, appears to
be much less disastrous than to ignore income eects and threshold levels.
Third, Post Keynesian consumer theory seems particularly relevant to
questions of environment, since environmental issues relative to material
wants seem a most obvious application of the principle of subordination
of needs (Gowdy and Mayumi, 2001).
M L
See also:
Agency; Consumption; Institutionalism; Environmental Economics; Non-ergodicity; Sraan
Economics; Uncertainty.
References
Drakopoulos, Stavros A. (1994), Hierarchical choice in economics, Journal of Economic
Surveys, 8 (2), 13353.
Earl, Peter E. (1986), Lifestyle Economics: Consumer Behaviour in a Turbulent World,
Brighton, UK: Wheatsheaf.
Eichner, Alfred S. (1987), The Macrodynamics of Advanced Market Economies, Armonk, NY:
M.E. Sharpe.
Gowdy, John M. and Kozo Mayumi (2001), Reformulating the foundations of consumer
choice theory and environmental valuation, Ecological Economics, 39 (2), 22337.
Gualerzi, Davide (1998), Economic change, choice and innovation in consumption, in
Marina Bianchi (ed.), The Active Consumer: Novelty and Surprise in Consumer Choice,
London: Routledge, pp. 4663.
Lancaster, Kelvin (1991), Hierarchies in goods-characteristics analysis, in Kelvin Lancaster,
Modern Consumer Theory, Aldershot, UK and Brookfield, VT, US: Edward Elgar, pp. 6980.
Lavoie, Marc (1994), A Post Keynesian theory of consumer choice, Journal of Post Keynesian
Economics, 16 (4), 53962.
Lutz, Mark A. and Kenneth Lux (1979), The Challenge of Humanistic Economics, Menlo
Park: Benjamin/Cumming.
Consumption
Accounting for 70 per cent or more of total spending on final goods and
services, consumption is by far the most important component of aggregate demand. The modern theory of aggregate consumption dates from
Keynes, who proposed that a propensity to consume governs the functional relationship between a given level of income and expenditures on
consumption out of that level of income. Keynes thought the propensity
to consume was a fairly stable function so that, as a rule, the amount of
Consumption
73
aggregate consumption mainly depends on the amount of aggregate

income. The propensity to consume itself is largely determined by a fundamental psychological law that consumer units (men originally) tend to
reduce their rate of consumption (C) as their income (Y) increases; that is,
dC/dY1 with CC(Y).
Keyness proposed relationship was quickly subjected to empirical
testing, using whatever data were available and whatever specification
seemed reasonable. Some anomalies were soon uncovered. First, estimates
of marginal propensities using cross-sectional household data ranging
from 0.4 to 0.8 could not be reconciled with those derived from time-series
aggregate data clustering around 0.9. Second, when transformed into an
average consumption model, the linear Keynesian model predicted a longrun declining spending rate not found in the actual data. Third, crosssectional functions showed shifting or ratcheting spending from one
dataset or year to another, behaviour not predicted by cross-sectional specifications or found with time-series data. Although the actual historical
development of these anomalies is more complex, they soon became a
fashionable feature of many macroeconomic textbooks and survey articles
(Thomas 1989, p. 131), generating an enormous literature and motivating
many resolutions.
Nearly all these resolutions focused on the consumption function
paradox, or dierences between estimates of cross-sectional and timesseries spending behaviour. The best known of these, the life-cycle theory of
Franco Modigliani and Richard Brumberg and the permanent income
hypothesis of Milton Friedman (reviewed in Deaton 1992; see also Carrroll
2001 and Muellbauer and Lattimore 1995), were essentially similar explanations developed from the neoclassical theory of household behaviour.
Like individual consumers, the aggregate or representative consumer unit
seeks to maximize the utility from lifetime consumption, an eort constrained by the present value of lifetime income plus any original endowments minus any bequests. Since consumption decisions are motivated by
lifetime considerations, the relevant income for these decisions is expected
lifetime or permanent income rather than current or transitory income.
While these distinctions are nebulous permanent to one is transitory to
another, in the context of aggregate spending average income was identified as permanent income and any other income as transitory. Thus, in the
neoclassical view, in its simplest form, when interest and time discount rates
are assumed to be zero, average consumption is a function of average
income. Since average data are time-series data, the appropriate marginal
propensities are the time-series ones, and since the Keynesian consumption
function utilizes current or cross-sectional data, it can be dismissed as
reflective of transitory, not permanent, spending behaviour.
74
Consumption
While simply a semantic exercise, the neoclassical reformulation of the

Keynesian consumption function seemed compelling. It was reasonable,
for obviously consumption is determined by more than current income. It
was grounded in the microfoundations underlying the education of all
economists. Its theoretical, mathematical and econometric complexities,
such as determining appropriate or tractable utility functions and measuring permanent income, oered research possibilities for the ambitious. By
having consumption determined by long-run considerations, it found
favour with anti-Keynesians. But, most importantly, proponents of the
Keynesian view oered no eective counter-arguments to the permanent/transitory distinction. As a consequence, macroeconomics shifted
away from the implications of current spending behaviour and the
Keynesian consumption function was relegated to honorific status, largely
serving to motivate extensions of the neoclassical consumption model.
In fact, it is the case for the Keynesian view that is compelling. First, the
alleged consumption function paradox simply reflects confusion about the
geometric implications of cross-sectional and time-series data. The usual
diagrammatic representation of the paradox is shown in Panel A of Figure
4, with the slope of the cross-sectional (CS) function less than that of the
time-series (TS) one. The functions cross at Y*, average income for either.
In the neoclassical view, income dierent from Y* is transitory. Hence, in
the absence of transitory income, all incomes equal average income and fall
on the TS function, and the CS function disappears.
C
C
TS
TS
CS1
Y1*
Y
Panel A
Figure 4
The consumption function
CS2
CS1
Y1*
Y2*
Panel B
Consumption
75
None the less, CS functions still exist for every time period; two are
shown in Panel B. Even if the transitory notion is correct, the slopes (marginal propensities) of the CS and TS functions must dier, because the
slope of the TS line connecting points on the CS lines cannot equal the
slope of either CS line.
Second, the alleged predictive failure of the Keynesian formulation
reflects the failure to distinguish carefully between cross-sectional and timeseries data. CS functions exist for every time period:
cti at bt yti, where i1, . . ., m consumer units.
When transformed into an average function, the average propensity to
consume converges on bt as CS income increases. Similarly, a TS function
exists covering all years:
Ct Yt, where t1, . . ., n years.
When transformed into an average function, the average propensity converges on as TS income increases. The objection that the Keynesian
model produced incorrect predictions about the average propensity is
simply the consequence of introducing time-series values into the crosssectional function. When appropriate income data are used, the CS function predicts a declining average propensity as income increases in any time
period while, following the conventional assumption of 0, the TS function predicts a constant propensity over time. Both these predictions are
consistent with the historical record.
Finally, the eects of the definitional relationship between crosssectional and time-series data have not been realized (Bunting 2001). Like
corporations, time-series data are soulless, having no existence beyond the
individual consumption and income observations from which they are constructed. But these individual data are the cross-sectional data. Thus, since
Ct
cti and Yt
yti, average consumption and income used for TS estimation are also CS averages. This implies that, with average data, the CS
function can be stated as:
Ct at btYt
and the TS marginal propensity is:
dC/dYdat /dYdbtYt/dY.
From the TS function, the TS marginal propensity is also:
dC/dY .
76
Consumption
Assuming constant CS marginal propensities,

dat /dYb.
In words, the CS and TS marginal propensities dier only because of shifts
in the cross-sectional functions.
Time-series coecients are determined by two factors: an induced component, reflecting cross-sectional consumption behaviour, and an autonomous
component, reflecting shifts in that behaviour. The induced component
simply indicates that time-series behaviour requires behaviour in every time
period, while the autonomous component is necessary for time-series data to
exist. With direct estimation, it is not possible to determine whether changes
in time-series behaviour are the result of actual behavioural changes or the
result of autonomous, unknown influences. Since the time-series coecient
is exactly defined by these two components, time-series behaviour is meaningless after eliminating autonomous and induced eects, the time-series
coecients are zero.
For their part, Post Keynesians have not shown much interest in expanding or extending Keyness basic ideas. Since the consumption function is
derived from individual data, it actually shows how the distribution of
income aects consumption: for example, the rate of spending declines as
income rises. This can also be shown by calculating consumption and
income shares, wi ci /
ci and vi yi /
yi. Because every consumer unit must
have its own consumption function, the CS marginal propensity is an
income-weighted average of the individual propensities, bt
bivi, implying
that changes in the distribution of income produced by alterations in the
business cycle or by public policy will change the overall marginal propensity. Autonomous influences are also important, yet these simply indicate
that unknown factors influence spending. Determining these factors has
obvious merit. Quite possibly they could be demographic and sociological
factors, suggesting that the consumption function has a much larger social
dimension than is commonly recognized.
The Keynesian view that current income largely determines current consumption is commonly attacked as myopic and unrealistic. Consumer units
spend and earn over their lifetimes; most are not fooled by unanticipated
changes in current income caused by fortuitous or adverse events. Instead
they plan consumption, schedule purchases, set retirement goals, save for
rainy days, and the like. It is this focus on individual behaviour that forms
the basis of the neoclassical critique of Keyness theory. As an explicit
theory of individual spending behaviour, this perspective is relevant for
those able to save, who constitute half or less of all consumer units. For this
group, spending could very well be based on lifetime considerations.
Credit rationing 77
However, for the other half or more, those unable to save, annual consumption is simply a question of annual survival and lifetimes are determined
year by year.
Unfortunately for the neoclassical critique and for those concerned
about myopic spending, regardless of form, Keynes developed a theory of
aggregate, not individual, consumption. In macroeconomics (as opposed
to microeconomics) individual consumption decisions are revealed in the
aggregate data. The motivation for these decisions, or how they fit into
some individual lifetime spending plan, is of no consequence. Instead, the
aggregate data reveal the annual relationship between consumption and
income. While incompletely understood, suering misrepresentation by its
critics and neglect by its adherents, this relationship none the less governs
the operation of any aggregate economic system.
D B
See also:
Eective Demand; Multiplier; Saving.
References
Bunting, D. (2001), Keynes law and its critics, Journal of Post Keynesian Economics, 24 (1),
14963.
Carroll, C.D. (2001), A theory of the consumption function, with and without liquidity constraints, Journal of Economic Perspectives, 15 (1), 2346.
Deaton, A. (1992), Understanding Consumption, New York: Oxford University Press.
Muellbauer, J. and R. Lattimore (1995), The consumption function: a theoretical and empirical overview, in M.H. Pesaran and M.R. Wickens (eds), Handbook of Applied
Econometrics: Macroeconomics, London: Blackwell, pp. 221311.
Thomas, J. (1989), The early history of the consumption function, in N. de Marchi and C.
Gilbert (eds), History and Methodology of Econometrics, Oxford: Clarendon Press,
pp. 13149.
Credit Rationing
The theory of credit rationing developed by Stiglitz and Weiss (1981) has
received much attention in the economic literature. However, this New
Keynesian approach assumes asymmetric information, in which there is a
precise probability distribution of returns from potential investment projects known by the borrower but not by the lender. In contrast, the Post
Keynesian approach to credit rationing is based on the assumption of
Keynesian uncertainty, a non-ergodic future about which both borrower
and lender simply do not know. In addition to uncertainty, the Post
Keynesian approach builds upon the following concepts:
78
Credit rationing
Borrowers risk and lenders risk In the General Theory, Keynes

(1936, p. 144) defined borrowers risk as due to doubts about actually
earning the prospective yield for which he hopes. Lenders risk was
related to either voluntary default by the borrower (moral hazard) or
involuntary default due to the disappointment of expectation.
Financial fragility As Hyman Minsky (1986, p. 213) has argued,
the successful functioning of an economy within an initially robust
financial structure will lead to a structure that becomes more fragile
as time elapses. This increase in financial fragility is likely to occur
during the expansion phase of the business cycle or over the course
of a series of relatively mild business cycles.
Endogenous development of expectations The endogenous development of financial fragility, particularly if accelerated by rising interest rates and falling profit rates (two typical developments near the
end of the business cycle peak), leads to a corresponding change in
bankers willingness to lend. This endogenous development of
expectations (Crotty 1994) is likely to lead to a reduction in bank
lending.
Dow builds upon these concepts to argue for a Post Keynesian theory of
credit rationing. She argues that this theory contradicts the horizontalist
interpretation of endogenous money: a theoretical case is made for amending the horizontalist position to allow for systemic credit rationing, referring particularly to the business cycle (Dow 1996, p. 498). It would appear
that the idea of endogenous money, if interpreted to mean a horizontal
supply curve such that bankers accommodate all demands for loans, would
indeed be in conflict with a concept of credit rationing in which bankers do
not accommodate demands for loans.
Dow argues that, over the course of the business cycle, financial institutions become increasingly less willing to lend. Building upon Minskys
(1975) discussion of borrowers and lenders risk, she asserts that the
demand for borrowed funds and the supply of borrowed funds are less
interest elastic the greater is the perceived borrowers risk and lenders risk,
respectively (Dow 1996, p. 500). A more sharply rising supply curve meets
a more sharply falling demand curve, with the result being a reduction in
the amount of credit extended.
These developments in the economic and financial systems result in a
reduced availability of credit, which Dow identifies with credit rationing.
She recognizes that there may be an issue in equating these two terms: It is
a matter of semantics whether or not the resulting availability of credit is
termed rationing (p. 499). None the less, she uses this analysis to make her
main point, which is that this reduction in credit availability is incompatible
Credit rationing 79
with a horizontalist view of endogenous money that assumes accommodating credit behaviour on the part of the banking system.
Wolfson, however, attempts to develop a framework to analyse credit
rationing that incorporates a horizontal endogenous money supply curve.
He defines credit rationing as any situation in which the bank refuses to
lend to a particular borrower, despite the borrowers willingness to pay a
higher interest rate (Wolfson 1996, p. 463). In his analysis, he uses the following additional Post Keynesian concepts:
The fringe of unsatisfied borrowers In the Treatise on Money,

Keynes (1930, vol. I, p. 212) referred to an unsatisfied fringe of borrowers who are refused credit. In this way, Keynes argued, the banks
can increase or decrease the volume of their loans (and thus investment) without there being necessarily any change in the level of
bank-rate [or] in the demand-schedule of borrowers. Keyness
concept clearly implies a banking system that does not accommodate
all demands for credit. Moreover, his reference to changing the
volume of loans in the context of both an unchanged interest rate
and an unchanged demand schedule would seem to imply a movement o the borrowers demand curve, rather than along it.
Asymmetric expectations Wolfson introduces this term to make the
point that the source of credit rationing is the existence of an uncertain future, one in which borrowers and lenders come to dierent
conclusions about future events. It is not necessary to rely on either
asymmetric information or even the assumption that lenders are less
risk-averse than borrowers. Just dierent, or asymmetric, expectations will lead to a situation in which borrowers will have some projects that they regard as risky, whereas lenders do not, while
borrowers will have other projects that they regard as safe, whereas
lenders do not. The first group of projects will never be seen by
lenders, while the second group will be rationed (refused credit).
Wolfson draws upon surveys conducted by the Federal Reserve (the

Survey on Bank Lending Practices and the Survey on the Terms of Bank
Lending) to understand how banks actually ration credit. He concludes
(1996, pp. 45960) that:
1.
Banks classify borrowers according to perceived risk, and use these risk
classifications to set both price and non-price terms of lending. (Price
terms refer to explicit charges, such as the cost of credit lines or the
spread of loan rates over a base rate; non-price terms involve other
aspects of the loan agreement, such as requirements for collateral, loan
covenants, or the size of credit lines.)
80
Credit rationing
2.
Higher price terms imply a higher spread over the reference rate, except
for borrowers with loan commitments (in the short run).
Higher non-price terms raise requirements on existing loans, but also
provide the basis for denying credit to those borrowers judged to be
insuciently creditworthy.
Banks give preference to borrowers with whom they are familiar.
3.
4.
These observations lead to an important conclusion: since banks

increase both price and non-price terms in response to perceptions of
higher risk, and since higher non-price terms are the basis for denying a
loan, spreads over reference rates and credit rationing move together. This
is interesting because, for those borrowers who are rationed, the bank has
denied them credit rather than increasing the interest rate charged.
Wolfson builds upon these observations to develop a framework for
understanding credit rationing. In doing so, he introduces two additional
concepts:
Notional demand curve The notional demand curve expresses the

desire for loans by borrowers. It is the traditional demand curve used
in most situations to analyse the interaction between supply and
demand for a commodity. However, the demand for bank loans is
dierent from the demand for most other commodities. To obtain a
bank loan, it is necessary to achieve the consent of the lender. In the
situation of credit rationing, it is precisely this consent that is not
given. Thus the borrowers demand curve is simply a notional (theoretical, not eective) demand curve for the bank.
Creditworthy demand curve What the bank uses is the creditworthy
demand curve. This represents the banks judgement about the proportion of borrowers desiring loans who are creditworthy. (In the
1996 article, Wolfson used the term eective demand curve, rather
than creditworthy demand curve. However, as pointed out by Peter
Skott and Marc Lavoie, use of the term eective demand is unwise,
since it already has a specific meaning within Keynesian economics.
Creditworthy demand is a better term (Wolfson 1997; see also
Rochon 1999).
Wolfson emphasizes that the judgements made by the bank are those
made in the context of uncertainty. Thus they are subject to two concepts
discussed by Keynes (1936): a state of confidence about any particular forecast of the future, and lending conventions that bankers seize upon in the
absence of any firm knowledge of the future. Both of these are subject to
change. They can change gradually and endogenously over the course of
Credit rationing 81
the business cycle, but they can also change dramatically and suddenly
during more turbulent times.
Wolfsons ideas can be illustrated with the use of Figure 5. Here L indicates bank loans, r the interest rate charged by the bank, DNDN the notional
demand curve, DCDC the creditworthy demand curve, C the banks cost of
funds, SA the average spread over the cost of funds charged to the borrower
(so that SA C represents the interest rate paid by the borrower), and ABD
the horizontal endogenous money supply curve. The bank accommodates
all creditworthy demands for credit and rations the rest. Thus BD represents the amount of credit rationing, or the fringe of unsatisfied borrowers who would be willing to pay a higher interest rate, but who are refused
credit.
r
DN
SB + C
SA + C
D
A
B
DN
DC
L
0
Figure 5
Wolfsons model of credit rationing
If the bank increases its average spread to SB, then the amount of credit
rationing would increase to FG. However, Wolfson indicates that the main
action in the model would have less to do with relatively small changes in
interest rates at a given point in time, and more to do with factors that aect
banks expectations of the future. He gives three examples: the boom and
bust lending in the 1970s and 1980s, the quick collapse of lending that
accompanies financial crises, and the typical change from optimism to pessimism (discussed by Minsky and Dow) that occurs as the business cycle
expansion nears its peak.
M H. W
82
Critical realism
See also:
Banking; Endogenous Money; Expectations; Financial Instability Hypothesis; New
Keynesian Economics; Non-ergodicity; Uncertainty.
References
Crotty, James (1994), Are Keynesian uncertainty and macrotheory compatible? Conventional
decision making, institutional structures, and conditional stability in Keynesian macromodels, in G. Dymski and R. Pollin (eds), New Perspectives in Monetary Macroeconomics,
Ann Arbor: University of Michigan Press, pp. 10539.
Dow, Sheila (1996), Horizontalism: a critique, Cambridge Journal of Economics, 20 (4),
497508.
Keynes, John Maynard (1930), A Treatise on Money, New York: Harcourt, Brace.
Keynes, John Maynard (1936), The General Theory of Employment, Interest and Money, New
York: Harcourt, Brace.
Minsky, Hyman P. (1975), John Maynard Keynes, New York: Columbia University Press.
Minsky, Hyman P. (1986), Stabilizing an Unstable Economy, New Haven: Yale University
Press.
Rochon, Louis-Philippe (1999), Credit, Money, and Production: An Alternative Post Keynesian
Approach, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Stiglitz, Joseph and Andrew Weiss (1981), Credit rationing in markets with imperfect information, American Economic Review, 71 (3), 393410.
Wolfson, Martin H. (1996), A Post Keynesian theory of credit rationing, Journal of Post
Wolfson, Martin H. (1997), Letter to Marc Lavoie, 12 March.
Critical Realism
Although known to economists through the work of Tony Lawson and
others since the late 1980s, critical realism has a long history and cannot
adequately be characterized by the work of any single author or by application within any one discipline. With that caveat key features are presented
below: (i) the basic tenets of critical realism, systematized by Roy Bhaskar,
which remain definitive of critical realism; (ii) a very brief indication of the
development of critical realism since its inception; (iii) the impact of critical realism in (especially Post Keynesian) economics.
Bhaskar (1975) articulates a philosophy of natural science that he terms
transcendental realism. This philosophy proposes that objects of science
cells, molecules, atoms, sub-atomic entities, and so on have an intrinsic
structure (real essence) and associated modes of activity (mechanisms).
Bhaskar criticizes the received view, enshrined in the covering-law model
of scientific explanation, that scientific laws refer to event regularities. He
argues, instead, that laws refer to the aforementioned mechanisms and that
only under conditions of experimental control does a mechanism necessarily produce an event regularity. Outside of experimental conditions a mechanism acts as an enduring tendency, interacting with other tendencies, to
produce the flux of events. In other words reality is an open system,
Critical realism
83
whereas the controlled environment of experiment creates a closed

system. Furthermore, if laws did refer to event regularities, then their application to reality would require that (atomistic) event regularities are ubiquitous outside of experiment; that is, it would imply that reality is a closed
system such that experiment is unnecessary. Bhaskar opposes such reductionism, which he sees as a legacy of positivism. He proposes, instead, that
reality is stratified. Higher strata, such as the molecular level (or, higher
still, the cellular level) are causally irreducible to the lower strata from which
they emerge.
Bhaskar (1979) also oers a philosophy of social science, termed critical
naturalism. He argues that social structures and agents are each emergent
strata, irreducible to the natural realm, or to each other. On Bhaskars view
social structures, defined as ensembles of social relations, are reproduced
(or transformed), often unintentionally, by agents. For example, by getting
married a couple unintentionally help to reproduce the institution of marriage; by going to work the worker unintentionally helps to reproduce the
social relation of wage labour/capital; by paying rent the tenant unintentionally helps to reproduce the social relation of landlord/tenant. Thus
social structures constrain and enable the very practices through which they
are reproduced. A problem for social science is that experiment is largely
impossible. The compensator for this inability to undertake experiment is
the social scientists preconception, gained through their ongoing social
activities, of social structures (however distorted it may be). To take one
example, agents must, in order to participate in the economy, have concepts
of money and capital. The social scientist (who is also an agent within
society) uses such preconceptions as premises for hypothesizing the deep
social structures that give rise to, or condition, agents preconceptions. This
explanatory move, from preconception to deep social structure, is termed
retroduction. Through the 1980s, Bhaskars ideas gained considerable currency. They were developed and debated by radical philosophers, sociologists and other social scientists, and became known as critical realism (a
combination of critical naturalism and transcendental realism).
Since the late 1980s, Tony Lawson, followed by a group of economists
originally based at Cambridge University, has given voice to critical realism
in economics. The most comprehensive statement of their common position (beyond the consolidation of the basic tenets noted above) is found in
Lawson (1997). Lawson has added significantly to the critical realist
method. Whereas Bhaskar argued simply that preconceptions provide the
key to social scientific retroduction, Lawson argues that the social world
gives rise to partial or demi- event regularities. It is such demi-regs,
rather than simple preconceptions, that, in general, enable retroduction,
according to Lawson. These demi-regs do not presuppose an atomistic
84
Critical realism
reality, as in positivism. Instead, they serve to direct social scientists attention to where underlying social structures and mechanisms may be located.
Lawson gives as an example the demi-regularity of the historically poor
comparative productivity performance of the UK. This demi-reg initiates
an investigation into the underlying social structures that may account for
it; Lawsons previous work argued that one such social structure is the
system of industrial relations.
Lawson has used critical realism to criticize mainstream economics. He
argues that the essence of mainstream economics is the extensive deployment of a method to which statements of the form whenever X then Y
(where X and Y refer to events or states of aairs) are central. To explain
something is to deduce it from axioms, assumptions and conditions that
require statements of the aforementioned form (this is a variant of the covering law model of explanation). Lawson criticizes this deductivist
method, and hence mainstream economics, according to Bhaskars general
critique of positivism: the method fails to acknowledge that social science
is concerned primarily with deep social structures and agents rather than
with event regularities. In attempting to close the system, the content of
mainstream economics is pushed towards a conceptual world of ubiquitous
event conjunctions among atomistic individuals. Yet the real social world is
an open system of social structures and agents, not a closed system of
atoms. Until deductivism is jettisoned, in favour of methods adequate to
open social reality, mainstream economics (both econometrics and economic theory) will remain broken-backed.
Lawson (1994) has contributed to ongoing debate regarding the coherence of Post Keynesian economics. He argues that most, perhaps all, key
Post Keynesian nominal manifestations, such as opposition to the mainstream; emphasis on making method explicit; focus on uncertainty and
history; upholding of genuine human choice; allowance of competing substantive perspectives; and association with certain classical economists, can
be rendered intelligible by critical realism. Accordingly, he suggests that
Post Keynesian economics is made coherent if grasped as being essentially
a critical realist project. Two significant corollaries are: (i) it would appear
to be dicult to distinguish Post Keynesian economics from much (old)
institutional, Austrian, Marxian and other radical economics, which are
likewise redolent of critical realism, in their acknowledgement of open
systems; and (ii) the neo-Ricardian commitment to the deductivist method
would appear to exclude it from Post Keynesianism (Pratten, not Lawson,
makes this claim explicit; see Fleetwood 1999).
Largely as a result of the Cambridge groups eorts (including the weekly
Cambridge Realist Workshop, ongoing since 1990, whose list of speakers
reads as a Whos Who of the economic methodology discipline), there has
Critical realism
85
been a remarkable rise to prominence of critical realism within: (i) nonmainstream economics, especially within Post Keynesianism; and (ii) the
discipline of economic methodology (itself a burgeoning discipline).
Taking, firstly, Post Keynesians, there has been a general, though cautious,
acceptance of critical realism by leading Post Keynesians. Philip Arestis
explicitly characterizes Post Keynesianism as critical realist; Sheila Dow
has suggested that her Babylonian method is largely compatible with critical realism. The more critical reactions from within Post Keynesianism,
notably Davidson (JPKE 1999), have tended to focus on issues of strategy
and style (critical realism certainly does contain awkward neologisms).
Walters and Young (JPKE 1999) argue for a rejection of the marriage of
critical realism and Post Keynesian economics but do so from outside of
Post Keynesianism. Of course, by no means every detail of the Cambridge
groups arguments is endorsed, and a major area of controversy concerns
econometrics and mathematics, where non-Cambridge-based critical realists have claimed, contra the perceived position of the Cambridge group,
that critical realism is compatible with econometrics and mathematical
modelling.
A number of criticisms have been levelled against critical realism (see, for
example, Fleetwood 1999; JPKE 1999). Lawsons replies (and the several
replies to Stephen Parsonss series of critiques by members of the Cambridge group) have been robust but there is clearly a range of areas that it
remains for critical realists to develop. From the above-mentioned collections, and elsewhere, the following criticisms can be noted. Critical realism:
(i) does not add much to economic methodology since economics does not
posit unobservable entities, beyond those that are trivial; (ii) undertakes a
misplaced critique because mainstream modelling does not deal in events
at all and hence cannot be deductivist; (iii) misleadingly assimilates very
dierent levels of analysis under the single rubric of retroduction; (iv)
employs an unhelpful notion of transcendental deduction in philosophy;
(v) falsely states that event regularities of interest to science seldom occur;
(vi) employs an opaque notion of social structure; (vii) puts forward a tautological and banal conception of social structure and agency; and (viii)
provides little methodological help to the social scientist.
From a critical realist perspective, many of these criticisms appear to
miss the main point of critical realism in social science: social science
should be concerned with ensembles of social relations and with their
reproduction or transformation by agents. Many critics do not oer anything like this persuasive description of economic reality, and their criticisms appear myopic as a result. They appear to maintain the philosophical
bias against ontology (the theory of being) that critical realism exposes.
Indeed, it is not the philosophical justification, but rather the appeal to
86
Critical realism
intuition, that has most likely attracted economists and others to critical
realism. While philosophical arguments are important, it is only once they
are allied to fruitful methodological and substantive arguments that they
are likely to make a real impact. In this vein, Brown (2001) and Brown et
al. (2002) reproduce some of the criticisms listed above, and make other
criticisms, as part of a positive agenda that embraces the intuitive appeal of
critical realism.
Critical realism has made much ground within economics. It seems that
it will continue to do so in the future, not only through the eorts of the
Cambridge group, but also through the many new converts to critical
realism who are now engaged in actively promoting and developing it.
A B
See also:
Babylonian Mode of Thought; Econometrics; Sraan Economics; Treatise on Probability.
References
Bhaskar, R. (1975), A Realist Theory of Science, Leeds: Leeds Books.
Bhaskar, R. (1979), The Possibility of Naturalism: A Philosophical Critique of the
Contemporary Human Sciences, Brighton: Harvester.
Brown, A. (2001), Developing realistic philosophy: from critical realism to materialist dialectics, in A. Brown, S. Fleetwood and J. Roberts (eds), Critical Realism and Marxism,
London: Routledge, chapter 9, pp. 16886.
Brown, A., G. Slater and D.A. Spencer (2002), Driven to abstraction? Critical realism and the
search for the inner connection of social phenomena, Cambridge Journal of Economics,
26 (6), 77388.
Fleetwood, S. (ed.) (1999), Critical Realism in Economics: Development and Debate, London
and New York: Routledge.
JPKE (1999), Journal of Post Keynesian Economics, 22 (1), symposium on critical realism.
Lawson, T. (1994), The nature of Post Keynesianism and its links to other traditions: a realist
Lawson, T. (1997), Economics and Reality, London and New York: Routledge.
Development Finance
Two contrasting features often characterize developing economies. On the
one hand, economic development is often associated with a significant
demand for resources to finance accumulation, technical change and
growth. On the other hand, the domestic financial structure, required to
channel these resources in appropriate conditions (in terms of maturity
structure and costs), is too often underdeveloped (if not non-existent).
In the development economics literature, these issues are addressed from
a macroeconomic as well as from an institutional perspective. From the
macroeconomic standpoint, the debate often focuses on the determinants
and the allocation of saving in developing economies. As usual in macroeconomics, there is a lively debate on the question of causality: some claim
that aggregate saving is a requirement for investment and growth, and
others argue that the causality is the other way round.
On top of the debate on the causality between aggregate saving and
investment, there is an additional issue concerning the relevance of dierent domestic financial systems in economic development. Even though few
economists nowadays doubt that the degree of financial development is an
important requirement for economic growth, there is an ongoing debate on
the policies required to achieve such a development.
Currently the literature on the issue is dominated by the ShawMcKinnon
and the New Keynesian approaches. In a nutshell, the first one (for a survey
see Agnor and Montiel 1996) claims that financial markets in less-developed
countries are underdeveloped due to the historical repression of financial
systems (through interest rate ceilings, directed credit and so on). Financial
repression would cause low saving and low financial deepening, and deter the
development of the financial system.
In turn, New Keynesian models focus on the availability and distribution
of information between borrowers, lenders and financial institutions. The
relaxation of the perfect information hypothesis within a Walrasian framework permits these models to show that, in the context of asymmetric information, credit (and equity) rationing is to be expected in any market
economy, and it tends to be more prominent in developing ones. In addition, if informational asymmetries introduce ineciencies in financial
markets, they may have quantitatively significant real eects (Gertler 1988,
p. 560). In addition, adverse selection implies that capital is being ineciently allocated. In other words, in these cases funds for investment will be
lower than their potential and the allocation of resources will be distorted.
87
88
Development nance
An alternative view on the issues related to the financing of economic

development is based on Keyness paradigm of the monetary production
economy and Minskys (1982) financial fragility hypothesis. The startingpoint of Keyness analysis is his rejection of Says Law through his theory
of eective demand and his identification of investment as the causa
causans in the determination of output and employment. This obviously
requires that investment finance be independent from previous saving
(Studart 1995).
In Keyness story, the aggregate supply of investment finance is mainly
determined by the banks willingness to actively create deposits and credit,
and not by savers preferences. Therefore banks hold a key position in the
transition from a lower to a higher scale of activity (Keynes 1937, p. 668)
an assumption that seems to be deeply rooted in his description of the
evolution of the banking system in the Treatise on Money.
It is noteworthy that in this story the expansion of bank credit is not an
anomaly which necessarily leads to Wicksellian cumulative disequilibria (in
the form of inflation and forced saving): it is one of the most important
means by which a monetary economy can grow. A malfunctioning banking
system is not one which disrupts the equilibrium between true saving and
investment, but one which causes the failure of an important part of the
credit system and therefore a reduction in the capacity of the entrepreneur
economy to advance purchasing power to investors willing to accumulate.
It is bank credit, not saving, which plays the crucial role in the financing
of investment. This would appear to leave no role for savings, but such is
far from being the case. The key to understanding such a hidden facet of
Keyness economics has to do with the risks (in a Knightian sense) of
financing assets with long-term maturity in inherently uncertain market
economies where the demand for liquidity is always high for a significant
proportion of wealth-holders. This is where the issue of funding comes into
the story.
Funding can be strictly defined as the process of transformation of
short-term into long-term liabilities. From a Post Keynesian viewpoint
funding is a key concept in the analysis of problems related to the financing of long-lived assets in an economy where finance is mainly provided by
the management of short-term liabilities (that is, bank deposits) (Davidson
1986).
Because of the structure of banks liabilities, such credit is either short
term, and borrowers risk will be rising; or, if the banks agree to finance
long-term positions, they will be accepting higher liquidity risks. In one way
or another, growth will be followed by an increase of what Minsky termed
systemic financial fragility.
In the KeynesMinsky story, funding takes place through the issuing of
Development nance
89
long-term securities, that is, in primary markets. The existence of such

markets depends on the proper functioning of secondary markets (where
old securities are bought and sold), which, in turn, relies on continuous
trading to provide the liquidity to otherwise illiquid assets. It is this provision of liquidity that makes long-term bonds and securities attractive to
savers who, as Davidson (1986) has rightly put it, are searching for safe
liquidity time-machines, and rarely wish to be locked in to holding an
asset for a long period of time.
In this sense, funding can be interpreted as a response to a menacing
increase in both borrowers and lenders risks (Keynes 1936, p. 144). Hence,
investment finance in a world of uncertainty is characteristically a twofold
process of finance and funding (Keynes 1937, p. 664). Thus the question of
funding has interrelated micro- and macroeconomic facets. From the
microeconomic perspective, entrepreneurs and bankers desire to fund their
long-term commitments on a stable basis because of uncertainty about the
prospective conditions of credit and levels of interest rates. From a macroeconomic viewpoint, funding and, therefore, financial markets also play a
role, which is seldom spelled out: the role of mitigating the increasing financial fragility inherent in a growing monetary economy. It is important to
stress that financial fragility in itself is not a constraint on growth, but it
may disrupt the process of expansion. This is especially true if an increase
in fragility causes a debt-deflation, an expression of the exhaustion of
financial arrangements that may lead to depression.
From the perspective implicit in the ShawMcKinnon approach, low
financial depth and retarded development are direct consequences of financial repression and therefore the policy to solve such a problem is the
liberalization of domestic financial markets in order to increase the availability of domestic saving. In addition, many defenders of this position
would argue that opening the capital account would facilitate access to
foreign saving.
The New Keynesian approach presents an embarrassing challenge to the
view that financial markets are ecient allocators of capital, and makes
way for interventionist views that are foreign to the liberal wave that has
dominated academe since the 1980s. However, it also leads to the ambiguous view that, were it not for the problems generated by imperfect information or other market failures, that role would be fully restored and the
allocative eciency of capital would prevail. Even if there is room for an
analysis of a defective institutional framework (that is, one which is far
from the stylized single competitive capital market), the stimulus for saving,
especially by maintaining real interest rates, could be prescribed as the
means to increase saving and investment. In addition, this ambiguity in the
analysis leads to a highly ideological and inevitably inconclusive debate on
90
Development nance
whether government failures are more prominent than market failures

(Jaramillo-Vallejo 1994), and thus whether any policy to mitigate market
failures is not likely to make things worse.
In contrast, the Post Keynesian approach rejects the two most important
pillars of the conventional approach to development financing: the priorsaving argument and the ecient capital market hypothesis. Financial
systems do have a fundamental role in the process of economic development if they are able to transform short-term assets that are demanded by
savers as forms of liquidity time-machines into sources of funding with an
appropriate maturity structure to finance dierent economic activities and
capital accumulation. This role assumes dierent possible forms according
to the institutional background behind the financial structure.
Appropriate mechanisms to finance and fund growth and accumulation
are required for sustainable development. However, nothing can guarantee
that development will lead to a capital-market-based financial structure.
Indeed, most developing (as well as developed) economies still have a bankbased financial structure, where the capital market is weak and firms
depend heavily on credit for raising finance beyond that available from
retained earnings.
From a Post Keynesian perspective, the lack or underdevelopment of
organized financial markets can have two destabilizing consequences for
development. First, if financial markets remain underdeveloped and
funding is not available, banks liquidity preference will be high and they
may refrain from expanding their lending activity when the demand for
loans is rising rapidly. Second, if finance is forthcoming to sustain growth,
the financial position of both firms and banks will become more fragile
(how rapidly depends on the rate of growth). Furthermore, even if they do
lend more in times of growth, banks will almost certainly prefer short-term
loans (to finance consumption, working capital and/or speculation) to
longer-term, and hence riskier, investment projects. Ceteris paribus, if
banks are still prepared to finance expansion despite the lack of appropriate mechanisms to fund investment, the indebtedness of the corporate
sector must increase. Growth will only be sustained if some investing firms
borrow short, hoping to repay by borrowing until their investment matures
and begins to produce additional cash inflows (using Minskys terminology,
more and more investors and financiers will adopt speculative, and even
Ponzi, strategies).
Because the weight of speculative finance tends to increase with the
acceleration of investment, credit-based systems are thus extremely vulnerable to changes in credit conditions (especially shifts in interest rates) in
times of growth. If the financing of long-lived assets is supplied mainly
through short-term renewable loans, a change in the rate of interest will
Development nance
91
represent a significant rise in firms financial expenditures; if firms try to

adjust by cutting other expenditures simultaneously, this may set in motion
a vicious circle of financial reactions which could reduce eective demand
even further.
The Post Keynesian approach thus leads to distinct policy recommendations for dierent time horizons. In the long run, market-enhancing policies are required if private mechanisms to finance and fund investment are
to evolve and there are significant experiences in both developed and
developing economies of successful market-enhancing economies.
While such mechanisms are not developed, private credit (and equity)
rationing is likely to be a pervasive problem of developing economies. This
gives support for directed credit policies, especially to developing sectors
that, due to their long maturity horizon (for example, long-term fixed
capital accumulation) and/or risk characteristics (for example, technologyrelated investments and small and medium-sized enterprises) are likely to
have little access to private financing. Finally, given that most developing
countries have a bank-based financial structure, low and stable interest
rates are an important requirement to avoid inherent financial fragility
associated with resulting maturity mismatches evolving into undesirable
processes of financial instability and crises.
R S
See also:
Credit Rationing; Finance Motive; Financial Instability Hypothesis; Keyness Treatise on
Money; Monetary Policy; New Keynesian Economics.
Bibliography
Agnor, P.R. and P.J. Montiel (1996), Development Macroeconomics, Princeton, NJ: Princeton
University Press.
Davidson, P. (1986), Finance, funding, saving and investment, Journal of Post Keynesian
Economics, 9 (1), 10110.
Gertler, M. (1988), Financial structure and aggregate economic activity: an overview, Journal
of Money, Credit, and Banking, 20 (3), 55987.
Jamarillo-Vallejo, J. (1994), Comment on The role of the State in financial markets,
Proceedings of the World Bank Annual Conference on Development Economics 1993,
Washington, DC: World Bank, pp. 538.
Keynes, J.M. (1936), The General Theory of Employment, Interest and Money, London:
Macmillan.
Keynes, J.M. (1937), The ex-ante theory of the rate of interest, Economic Journal, 47 (188),
6639.
Minsky, H.P. (1982), The financial-instability hypothesis: capitalist processes and the behavior of the economy, in C.P. Kindleberger and J.P. Laargue (eds), Financial Crises,
Cambridge: Cambridge University Press, pp. 1339.
Stiglitz, J. (1994), The role of the State in financial markets, Proceedings of the World Bank
Annual Conference on Development Economics 1993, Washington, DC: World Bank,
pp. 1952.
Studart, R. (1995), Investment Finance in Economic Development, London: Routledge.
92
Dynamics
Dynamics
In analysing growth dynamics, Post Keynesian economics encounters a
series of deep contradictions that generate a startling array of alternative
models and approaches. These contradictions arise from confronting real
contradictions in actual growing economies that are papered over or
ignored by standard neoclassical approaches. In variations on the
SolowSwan model, the latter postulate long-run equilibrium growth paths
that are set by exogenous factors such as intertemporal time preference
rates, population growth, and proclivities for technological change, with
these characteristics essentially holding even in the newer, so-called endogenous growth models by Paul Romer and others. The essential contradiction for Post Keynesians arises from Keyness famous dictum that in the
long run we are all dead, and from Joan Robinsons equally famous
contrasting of logical versus historical time, while at the same time neoclassical growth theory arose out of long-run models by such followers of
Keynes as Roy Harrod, Nicholas Kaldor, and even Joan Robinson herself,
as well as the neo-Ricardian followers of Piero Sraa.
Thus more recent Post Keynesians attempt to overcome this contradiction by studying long-run dynamics without positing a long-run solution.
They seek to place these outcomes as the consequence of a sequence of
short-run solutions. In turn, these short-run solutions attempt to more
clearly integrate micro-sectoral outcomes with macrodynamic outcomes.
Relative change and transformation over time become important, with the
recognition that divergences, both internationally and internally between
groups, give rise to outcomes of interest. Furthermore, there is a much
greater emphasis on modelling the interaction between cyclical dynamics
and complex growth dynamics, with money and disequilibrium outcomes
playing much greater roles than in more orthodox approaches. We shall
review some of the approaches that have been developed along these lines.
Michal- Kalecki was the independent and parallel developer of the
Keynesian apparatus who drew on Marxist roots. More influential on many
modern Post Keynesians than Keynes himself, Kalecki introduced a more
clearly articulated view of microeconomic behaviour and also more clearly
focused on longer-run growth issues as well as shorter-run fluctuations
questions. All of this makes him in many ways paradigmatic of the more
general approach of Post Keynesians to these issues. Some of his more
important followers include Josef Steindl, Alfred Eichner, Joseph Halevi,
Peter Kriesler, Marc Lavoie, Tracy Mott and Malcolm Sawyer. An overview
of Kaleckian models is in King (1996).
At the micro level Kalecki emphasized monopoly power (as did Joan
Robinson also) and introduced the idea of mark-up pricing. This contrasts
Dynamics
93
with the usual marginal cost pricing theory of theoretical pure competition,
and is also known to correspond more closely with what actually goes on
in many real-world industries. This approach also allows for cost-push
sources of inflation, in contrast to the usual demand-side factors emphasized by monetarists or New Keynesians.
Growth is driven by capital investment, a view shared with the neoclassicals. But, in contrast to them, investment is endogenous to aggregate
demand, with savings endogenously determined by investment, a widely
held view among Post Keynesians that has considerable empirical support.
But fluctuations in investment, arising from fluctuations of aggregate
demand as well as from class struggle that sometimes operates through the
political arena, drive shorter-term fluctuations. Thus Kalecki was one of
the first to point the way to an integrated Post Keynesian approach.
Piero Sraa was a close ally of Keynes in his debates with Friedrich
Hayek about business cycles in the 1920s and 1930s, and Joan Robinson
was a close associate and follower of both men. Nevertheless, in recent
decades a deep divide has opened between Post Keynesians who emphasize
the monetary short-run equilibrium, such as Paul Davidson, and Sraans
or neo-Ricardians, who emphasize classical long-run supply-side equilibrium models based on inputoutput matrices. Leaders of the latter school
have included Piero Garegnani, Heinz Kurz and Neri Salvadori.
However, several economists have attempted to integrate the two
approaches, including Edward Nell and especially Luigi Pasinetti (1993).
Pasinetti became well known in the 1960s as a participant in the Cambridge
capital theory controversies, along with Garegnani, as one of the analysts
of reswitching. He also developed growth models based on capitalist saving
behaviour. He extended the Sraa inputoutput framework to a growth
context in which technical coecients could change. In the 1970s he altered
this by considering vertically integrated units within the inputoutput
framework. This opened the door to a more concentrated analysis of structural transformation over time.
In his more recent formulation, Pasinetti (1993) further decomposes the
analysis to a quasi-Marxist model based purely on labour, with no intermediate goods or capital (a logical outcome of the Sraa critique). This
version contains the earlier models, but has now reintegrated Keynesian
eective demand as an overall growth determinant. Furthermore, this
demand is sectorally specified and evolves through consumer learning.
Thus, although it is dierent in various ways from the Kaleckian formulations, Pasinettis model also integrates micro with macro in a model with
demand-determined growth.
The emphasis on relative sectoral transformation that one finds in
Pasinetti and Nell has also been carried forth by others who have brought
94
Dynamics
in the path-dependence dynamics of cumulative causation and hysteresis.

Such ideas appeared in Adam Smith and Alfred Marshall to some extent,
with Allyn Young restating them in the 1920s, and Nicholas Kaldor emphasizing their significance for the existence of multiple equilibria or no equilibria from the 1930s through the 1980s, and Paul David and Brian Arthur
emphasizing path dependence in the 1980s. Some of those following this
approach include Mark Setterfield and Rod Cross. But perhaps the most
comprehensive integration of these ideas into sectorally specific macro
growth models with sectoral and national divergence in specific historical
analysis is due to John and Wendy Cornwall (2001).
Uneven growth is more clearly analysed, with economies of scale eects,
learning eects and various nonlinearities playing important roles. Transformation is evolutionary, with divergences at the sectoral level ultimately
implying divergences at the macro level across nations and over time within
nations. The nonlinearities involved are recognized to possibly lead to
complex dynamics along the growth path (Rosser 2000). This contrasts
with the neoclassical endogenous growth models that remain at the level of
aggregate production functions, with all the diculties this entails, including a distribution theory based on marginal products of aggregated factors
of production. The Cornwall and Cornwall analysis, along with related
models by Setterfield and Nell, avoids these problems and is more revealing of historical economic dynamics of rising income inequality, both
internally and internationally.
An important contribution by Post Keynesians to the analysis of international elements in growth divergence has come from recognizing the role
of balance of payments constraints on economic growth. This has been
crystallized in what is known as Thirlwalls Law (McCombie and Thirlwall
1994). This posits that over time there must be something like a balance of
trade. Thus the growth of imports must be matched by the growth of
exports. This implies that a nations growth will be constrained by global
growth along with its import and export elasticities, the variables that enter
into the formula that expresses Thirlwalls Law.
Recognition of this has led to much discussion of the role of international finance more broadly in influencing economic growth and development. Post Keynesians who have discussed this at length and made specific
proposals include Paul Davidson, Jan Kregel and John Smithin.
Many Post Keynesians have especially stressed the importance of the role
of money in economic growth and fluctuations, thereby denying the classical dichotomy, with Paul Davidson and Hyman Minsky particularly important in this regard. However, they have not developed explicit growth models
along with their analysis of fluctuations. This has fallen to a more recent
group of economists, including Duncan Foley, Peter Skott, Domenico Delli
Dynamics
95
Gatti and Mauro Gallegatti, Steve Keen, Reiner Franke and Willi Semmler,
with Carl Chiarella and Peter Flaschel (2000) providing thorough modelling.
These models are more strictly macro than the ones discussed in earlier
sections, and also more clearly lay out the role of the financial sector in
investment, with feedbacks from production to finance along Minskyian
and Goodwinian lines. Growth interacts with cycles, with the latter becoming dynamically complex. Complex macro cycles have been studied by
many, including Richard Day, Alfredo Medio, Tnu Puu, and Rosser
(2000).
Finally, extending the analysis of Cornwall and Cornwall, there has
developed an essentially neo-Schumpeterian analysis of technical change
based on sectors that can lead to structural evolution and transformation
over time with complex dynamics, including historical reswitching. Important figures in this analysis have included Richard Day, John Sterman and
Gerald Silverberg. A general overview of these models and their complex
nonlinear dynamics that links them back to the Cambridge capital theory
debates and the historical versus logical time arguments of Joan Robinson
can be found in Rosser (2000).
J. B R, J.
See also:
Balance-of-payments-constrained Economic Growth; Growth Theory; Innovation;
Investment; Kaleckian Economics; Sraan Economics; Time in Economic Theory.
References
Chiarella, C. and P. Flaschel (2000), The Dynamics of Keynesian Monetary Growth:
Macrofoundations, Cambridge: Cambridge University Press.
Cornwall, J. and W. Cornwall (2001), Capitalist Development in the Twentieth Century: An
EvolutionaryKeynesian Analysis, Cambridge: Cambridge University Press.
King, J.E. (ed.) (1996), An Alternative Macroeconomic Theory: The Kaleckian Model and PostKeynesian Economics, Boston: Kluwer Academic Publishers.
McCombie, J.S.L. and A.P. Thirlwall (1994), Economic Growth and the Balance of Payments
Constraint, London: Macmillan.
Pasinetti, L.L. (1993), Structural Economic Dynamics. A Theory of the Economic
Consequences of Human Learning, Cambridge: Cambridge University Press.
Rosser, J.B., Jr. (2000), From Catastrophe to Chaos: A General Theory of Economic
Discontinuities, Volume 1: Mathematics, Microeconomics, Macroeconomics, and Finance,
2nd edition, Boston: Kluwer.
Econometrics
Perhaps emphasizing theoretical rather than methodological dierences
with neoclassical and new Keynesian economists, from the outset Post
Keynesian economists have made use of econometric methods. Thus in the
1980s Alfred Eichner with various research partners presented results associated with estimating blocks of structural equations for a short-period
Post Keynesian model of the US economy. Likewise, Philip Arestis undertook a comparable exercise for the UK. In general, casual perusal of the
Journal of Post Keynesian Economics would suggest that Post Keynesian
economists readily use a wide variety of econometric methods. Increasingly,
however, while Post Keynesian economics does not draw upon Keyness
ideas only, in the case of econometrics his somewhat ambiguous sentiments
are strongly echoed in the literature. Keyness essentially philosophical argument, that the main prima facie objection to the application of the method
of multiple correlation to complex economic problems lies in the apparent
lack of any adequate degree of uniformity in the environment (Keynes
1939, p. 567) is often cited. This said, Keynes (1973b, p. 300) also stresses
the need for messy acquaintance with the facts. So too, in the current Post
Keynesian literature, there is a degree of tension between philosophical pronouncements and the practice of economics.
Post Keynesian economics currently presents itself as accepting an
open-system philosophical approach from three main perspectives: critical realism associated with Tony Lawson, a Babylonian perspective associated with Sheila Dow, and an encompassing approach associated with
Paul Davidson. Critical realism in economics is primarily associated with
the work of Lawson (1997), who argues that neoclassical economics has
its roots in the philosophical system of positivism and in particular
embraces an ontology where reality comprises the constant conjunction
of atomistic events in a closed system. Here, broadly speaking, the intrinsic condition of closure (ICC) that each cause produces the same eect
and the extrinsic condition of closure (ECC) that each eect has the same
cause allow an epistemology based on deduction. Thus theoretical explanation can comprise statements of the form whenever event X then event
Y, allowing also for stochastic errors. Consequently the mathematical
modelling of individual agency is emphasized in neoclassical economics,
coupled with econometric testing. Lawson describes this approach as
empirical realism.
In contrast, an open-system approach presents an organic ontological
96
Econometrics
97
perspective, which implies that human agency is embedded in a social

context. Behaviour is thus irreducible to individual action per se but, on the
contrary, is both conditional on, and results in, multiple modes of the
determination of events. Furthermore, critical realists argue that reality is
stratified into three domains. These are the level of actual events, the empirical level of experience and sense impression, and the level of the real, where
causal relations are located. Accordingly, critical realism maintains that at
best there will be a plurality of partial regularities and processes underlying events, and not predictable or universal event-regularities. Econometric
inferences are thus inherently problematic.
Davidsons (1996) methodological approach broadly shares these sentiments. He argues that, in general, the neoclassical research programme
invokes the axiom of ergodicity. This implies that the world is predetermined and immutable. In the case of probabilistic inferences, therefore, as
the past is a good guide to the future, objective or subjective probabilities
will ultimately converge on the true values of the parameters of the probability distribution. For Davidson, the ergodicity axiom is the reason why
neoclassical economists emphasize probabilities and statistical/econometric inference in their analysis. In contrast Davidson argues that Post
Keynesians embrace a non-ergodic and transmutable-reality view of the
world in which probabilities, and thus econometric inference, are not reliable guides to the future.
Finally, while echoing the critical realist perspective in stressing the
organic nature of society, Dows (1990) Babylonian approach argues that
evidence is validly provided by a variety of sources, such as questionnaire
and historical sources. This approach also allows for qualified econometric
testing, however, because theories cannot be judged according to the principles of a particular theoretical structure and because theoretical and
empirical diversity is a logical consequence of open-system thought.
For the purpose of further discussion, though adopting the particular
language of critical realism, Post Keynesian philosophical deliberation in
general rejects empirical realism. However, the practice of Post Keynesian
economics admits of the need for empirical analysis. Herein lies the central
tension of this discussion. It is clear that both the estimation of regression
coecients and an emphasis upon drawing statistical inferences require the
invocation of the closure conditions noted earlier. The ICC is equivalent to
assuming the underlying homogeneity of nature and the atomistic combination of objects. This is required to ensure that the coecients, or functional form, of a regression are constant over time (or space). The ECC
implies that all of the causal factors have been included in an econometric
study, or that the eect of external factors on internal factors is constant.
This last point is equivalent to assuming that countervailing factors are
98
Econometrics
constant. As noted above, the fact that Post Keynesians, like mainstream
economists, regularly employ econometric methods, suggests a possibility
of logical inconsistency between the methodological aspirations and the
practices of Post Keynesians.
In fact, though not clearly articulated in the literature, the employment
of econometrics could also reflect the need for Post Keynesians to adopt an
appropriate rhetoric in order to engage in critical discussion with neoclassical economists. It has often been argued that Post Keynesian economics
only has coherence in terms of an opposition to neoclassical economics. In
a related way econometrics might also act as a vehicle for them to demonstrate comparable technical virtuosity with neoclassical economists.
Yet these are rather unsatisfactory and defensive arguments. Post
Keynesians have also sought to present a positive case for employing
econometric methods that confronts the issues raised by Post Keynesian
philosophical concerns. Summarizing a number of arguments, Downward
and Mearman (2002) explore the variety of econometric methods,
arguing that not only do a burgeoning set of techniques exist for the
researcher to calculate statistics, but that they also form part of various
inferential frameworks.
Thus, for example, the average economic regression or textbook
approach presents a maintained hypothesis that is assumed to capture a
correct specification. Following estimation, the random error terms are
analysed and transformations of the model then follow to eliminate any
problems. In contrast, the David Hendry/LSE (London School of Economics) approach stems from developments in time-series econometrics.
Central to Hendrys research programme is a continual interaction between
theory and data; thus knowledge appears to arrive from a complex interaction of deduction and induction. This is demonstrated, for example, in the
focus upon error-correction models in the cointegration analysis of time
series. Theory describes the long-run relationship, while the data reveal the
short-run dynamics of adjustment.
Moreover, in the former case, it can be argued that the average economic
regression approach has its roots in Haavelmos (1944) influential essay.
For Haavelmo the essence of econometrics was to build, identify, estimate
and assess various models conforming to the optimizing behaviour given
by neoclassical precepts. Importantly, he writes, the question is not
whether probabilities exist or not, but whether if we proceed as if they
existed we are able to make statements about real phenomena that are
correct for practical purposes (Haavelmo 1944, p. 43). On such a basis
econometrics can be described as a form of instrumentalist reasoning. In
contrast, at times Hendry argues that the literal process that generates data
the data-generating process can be measured by probabilities as a sta-
Econometrics
99
tistical generating mechanism. However, he also stresses that the proof of

the pudding lies in the eating as far as econometric models are concerned.
Other econometric methods draw upon Bayesian inferential logic or
atheoretical vector-autoregression analysis. In the former case Ed Leamers
modus operandi is to formulate a general family of models, decide what inferences are of importance, which need not be structural relations, express these
in terms of parameters, and form prior distributions summarizing any information not in the data set. The sensitivity of inferences to a particular choice
of distributions should be analysed to explore their fragility. In the latter
case the concept of exogenous variables is rejected and analysis proceeds in
terms of simultaneous relationships between jointly endogenous variables.
Finally, there exists an eclectic set of broadly cross-sectional methods. For
example, Logit and Probit estimators are employed in cases in which the
dependent variable is presumed to reflect simple dichotomous categories, or
rank-orders. Poisson and negative binomial models are now estimated for
models which maintain that dependent variables reflect discrete rather than
continuous values (that is, comprise integers or counts). Other developments
include allowing for truncated or censored distributions in the dependent
variable. The most famous example of this is the Tobit model. Under these
circumstances, the assumption is usually that the dependent variable is
drawn from a normal distribution but that there is a qualitative break in the
measurement of the variable. Interestingly, these latter approaches indicate
that econometricians have thought quite deeply about ontological issues.
Estimators have been refined in connection with presuppositions about the
purported character of phenomena.
Despite these dierences, however, Downward and Mearman argue that
the underlying inferential logic of most econometric techniques, as typically applied, appeals to measurable probabilities alone. It follows that
embracing current philosophical discourse suggests that Post Keynesian
inferences cannot simply proceed by appeal to a measurable probability distribution (and by implication a unique set of estimated coecients).
However, Downward and Mearman also argue that the assumptions
involved in estimating coecients are shared with all (even descriptive)
empirical analyses. If this is so, then logically any empirical analysis advocated by Post Keynesians can embrace econometric estimation. As far as
inferences are concerned, Keynes emphasized the importance of rational
belief rather than knowledge as a basis of argument (Keynes 1973a, p. 10).
To avoid the problem of induction, Keynes argued that one should examine
a particular phenomenon in dierent contexts, thus engaging in a process
of negative analogy. If a phenomenon appears to be a common element
between various contexts, then ultimately this can add weight to a particular account of that phenomenon.
100
Econometrics
From an operational perspective this suggests that various empirical

insights should be triangulated, that is, compared to insights produced
elsewhere. This idea has resonance with, or and can be combined with, the
tenets of critical realism. Critical realism strongly argues that analysis
should pay explicit attention to ontology and, in particular, explanation
should focus, by a process of retroduction, on elaborating the underlying
causal mechanisms of events. Crucially, these are not likely to be synchronized with empirical statements about the events. Consequently, Downward
and Mearman argue that while descriptive and historical analysis might be
employed to explore suggested causal mechanisms, it follows that the
eects of their action can be assessed, and hence the purported causal
mechanism supported, with reference to more quantitative analysis conducted by econometric techniques. Thus econometric methods can potentially perform a very helpful task in codifying events at the empirical level,
suggesting issues for further causal investigation and helping to assess the
legitimacy of existing causal claims.
It should be clear from this that Post Keynesians do not accept that
explanation and prediction are synonymous, as implied by much neoclassical presentation of econometrics. In Post Keynesian thinking, any quantitative prediction becomes merely a scenario whose legitimacy will rest
upon the robustness of the claimed causal mechanism that is tentatively
identified and/or supported by econometric estimation. Predictions from
an econometric model will always be open to revision.
In closing it should be emphasized that philosophy and applied economics need to coexist and to develop from mutual discourse, in terms of both
articulating the problems of adequately capturing real elements of economic processes and working towards concrete analysis and policy prescription. Some compromise with a purely philosophical inclination seems
inevitable in applying econometric, and indeed other empirical, methods.
P D
See also:
Agency; Babylonian Mode of Thought; Critical Realism; Non-ergodicity; Treatise on
Probability.
References
Davidson, P. (1996), Reality and economic theory, Journal of Post Keynesian Economics, 18
(4), 479508.
Dow, S.C. (1990), Post Keynesianism as political economy: a methodological discussion,
Review of Political Economy, 2 (3), 34558.
Downward, P.M. and A. Mearman (2002), Critical realism and econometrics: constructive
dialogue with Post Keynesian economics, Metroeconomica, 53 (4), 391415.
Haavelmo, T. (1944), The probability approach in econometrics, Econometrica (Supplement), 1118.
Economic policy 101

Keynes, J.M. (1939), Professor Tinbergens method, Economic Journal, 44 (195), 55568.
Keynes, J.M. (1973a), The Collected Writings of John Maynard Keynes. Volume VIII: A
Treatise on Probability, London: Macmillan for the Royal Economic Society.
Keynes, J.M. (1973b) The Collected Writings of John Maynard Keynes. Volume XIV: The
General Theory and After. Part II: Defence and Development, London: Macmillan for the
Lawson T. (1997), Economics and Reality, London and New York: Routledge.
Economic Policy
There is no unique set of economic policies which can be described as Post
Keynesian or even closely associated with one of the branches of Post
Keynesian economics. Post Keynesian economists have though generally
been supportive of certain types of economic policies, notably those to
stimulate the level of aggregate demand. Many policy proposals find
support from a range of economists of dierent schools of thought, and
there are many dierences over policy among Post Keynesian economists
(the proposals for a tax on foreign exchange transactions, often labelled the
Tobin tax, being a current example).
The general approach to policy could be seen to be derived from the basic
insight of Kalecki and Keynes for the workings of industrialized market
economies, namely that a laissez-faire market economy will not usually generate full employment. The essential cause of that failure to create full
employment is not some rigidities or imperfections of monopolistic competition, trade unions and so on which could potentially be removed
through government action. It is rather that a laissez-faire market economy
would exhibit elements of instability with booms and busts, and periods
of crisis. Further, a market economy would not usually generate a level of
aggregate demand consistent with full employment. The achievement of
full employment is a widely accepted major policy objective for Post
Keynesian economists, and policies and institutional arrangements supportive of high levels of demand are advocated.
The Post Keynesian approach to fiscal policy is informed by the accounting identity of:
(Savings Investment)(Exports Imports)
(Government expenditure Taxation),
where (Exports Imports) is the trade surplus, and equal to the deficit on
capital account. There will, in general, be imbalances in the terms of this
equation at whatever level of income the economy is operating. The question arises as to what the imbalances would be if income was at a level
102
Economic policy
compatible with full employment. In the case where savings would exceed
investment at full employment income, then some combination of trade
surplus and government budget deficit would be required to balance the
excess of savings over investment to maintain full employment. Running a
budget deficit in these circumstances helps to sustain full employment, but
it can also be seen that the excess of savings over investment funds the
budget deficit. Hence there is no crowding out, or upward pressure on
interest rates.
Lerner (1943) put the case for what he termed functional finance, which
rejects completely the traditional doctrines of sound finance and the
principle of trying to balance the budget over a solar year or any other arbitrary period (p. 355), and adjustment of total spending to eliminate both
unemployment and inflation. No matter how much interest has to be paid
on the debt, taxation must not be applied unless it is necessary to keep
spending down to prevent inflation. The interest can be paid by borrowing
still more (p. 356). Lerner summarized the answers to arguments against
deficit spending by saying that the national debt does not have to keep on
increasing, and that even if it does the interest does not have to be paid from
current taxes. Further, interest payments on bonds are an internal transfer.
From an aggregate demand stance, alternative policy approaches would
be the stimulation of investment or the reduction of the propensity to save.
For example, Kalecki (1944) in his discussion of three ways to full employment considered the redistribution of income (towards wage earners) as a
means of stimulating consumer demand (and thereby reducing savings). He
also considered the stimulation of investment, though he saw clear limits to
this route. Consider I/Y(K/K).(K/Y) where I is net investment equal to
the change in the capital stock K, and K/Y is the capitaloutput ratio. The
share of investment in GDP is then given by the multiple of the underlying
growth rate (which sets the growth of the capital stock K/K) and the
capitaloutput ratio.
Some Post Keynesian authors (for example, Wray 1998) have advocated
the policy of employer of last resort (ELR) whereby the government
stands ready to employ anyone at a pre-determined money wage (which
may be set to ensure that the wage is sucient to remove the worker from
poverty). It is argued that such a policy would not be inflationary (since the
wage paid by the government as ELR remains unchanged) and secures full
employment (in that anyone who wishes to work is able to do so, albeit at
the ELR wage).
There would be a concern to ensure that the international financial
system was conducive to high levels of aggregate demand. In the context of
the Bretton Woods fixed exchange rate system, this concern was translated
into policy arrangements which sought to ensure that countries which were
Economic policy 103

running trade deficits were not forced into deflationary policies to correct
the trade deficit. Keynes (1980, p. 176) sought to design an international
payments system which transferred the onus of adjustment from the debtor
to the creditor position and aimed at the substitution of an expansionist,
in place of contractionist, pressure on world trade. Davidson (1992), building on the work of Keynes, proposed a new international payments system
(p. 157) designed to resolve payments imbalances while simultaneously
promoting full employment economic growth and a long-run stable international standard of value (p. 153). These proposals included the unit of
account and reserve asset for international liquidity being the international
money clearing unit (IMCU), with each nations central bank committed to
guarantee one-way convertibility from IMCU deposits at the clearing union
to its domestic money. Davidson advocated an overdraft system to make
available short-term unused creditor balances at the clearing house to
finance the productive international transactions of others who need shortterm credit, and a trigger mechanism to encourage any creditor national
to spend . . . excessive credit balances accumulated by running current
account surpluses (p. 160). Finally, if a country is at full employment and
still has a tendency toward persistent international deficits on its current
account, then this is prima facie evidence that it does not possess the productive capacity to maintain its current standard of living (p. 163, original
emphasis).
The era of floating exchange rates has led to a dierent set of concerns,
namely the eects of the large flows across the currency exchanges and the
volatility of exchange rates. The volatility of exchange rates (and indeed of
prices in financial markets more generally) comes as no surprise to Post
Keynesians. In a world of uncertainty, where the equilibrium price in a
market is unknowable (and indeed may not exist in any meaningful sense),
traders will not (and cannot) hold rational expectations of future prices.
They will be influenced by the views of others, and by a variety of information, including recent trends in price. One policy proposal which addresses
these concerns is the Tobin tax.
Inflation, or the fear of inflation, may be seen as limiting the achievement
of full employment, though many Post Keynesians would cast doubt on the
existence of any immutable natural rate of unemployment or deny that
any such rate is eectively determined in the labour market. The Post
Keynesian views on endogenous money would deny any causal role for
money in the inflationary process. Deflation (whether brought through
fiscal or monetary policies) is seen as a blunt and inecient instrument for
the control of inflation. Some would envisage that any inflationary problems could be addressed through incomes policy of the social contract form
or through tax-based incomes policy, while others would point to the
104
Economic policy
importance of institutional arrangements and the creation of sucient

productive capacity.
The attainment of high levels of productivity requires an appropriate
work intensity and commitment on the part of the labour force as well as
the provision of sucient capital equipment, training, skills and management. In many market economies, unemployment (and more particularly
the threat of it) serves as a significant mechanism for imposing a high level
of work intensity (Kalecki 1943). Unemployment is seen as performing a
systemic function (of aiding the disciplining of workers). This is not to
argue that unemployment is necessary to ensure work eort; indeed unemployment heightens fear and brings demoralization, which serve to undermine it, and dierent market economies have drawn on mechanisms other
than unemployment.
The factors which influence investment and thereby the size of the capital
stock (such as profitability and capacity utilization) are generally rather
dierent from the factors which determine the size of the work force
(mainly demographic). From that crude observation, we could say that
there is no particular reason to think that the capital stock will be adequate
for the provision of full employment. There are rather limited opportunities, especially in the short run, for substitution between labour and
capital. We would expect that after a period of prolonged slow growth
(such as the past two decades) investment may have fallen short of what
would be required to sustain full employment.
The pace of growth is generally seen as driven by the growth of demand,
with supply (of labour, capital equipment and so on) adjusting to the
growth of demand. The eective supply of labour can vary through
changes in labour force participation rates, training and skill acquisition
and movement of workers from areas of disguised unemployment and low
productivity. The amount of capital equipment is determined through the
cumulative eect of investment.
The foreign trade position can constrain the rate of growth, in that a
growing deficit will emerge if there is a tendency for the growth of imports
to exceed the growth of exports. When the (domestic) income elasticity of
demand for imports is greater than the (world) income elasticity of demand
for the countrys exports, then the maintenance of a non-exploding trade
deficit requires that the domestic growth rate is suciently below the world
growth rate so that actual imports and exports grow in line with one
another (McCombie and Thirlwall 1997). One policy implication which
can be drawn is that continuing growth of output requires growth of
exports, and supply-side policies (such as industrial policy) can be required
to ensure the production of goods and services for which there is a strong
export demand.
Eective demand 105

The general Post Keynesian approach would be based on the view that
the creation of high levels of aggregate demand is necessary for the achievement of full employment. The creation of high levels of aggregate demand
can proceed directly through higher government expenditure and/or lower
taxation, indirectly through the stimulation of investment and the redistribution of income towards higher spending groups, and through the encouragement of institutional arrangements which are conducive to high levels
of demand. But a high level of aggregate demand is a necessary, though not
sucient, condition. Policies to ensure adequate capacity, the creation of a
low-inflation environment, an equitable distribution of productive activity
and a sustainable balance of trade position are also required (Arestis and
Sawyer 1998).
M S
See also:
Balance-of-payments-constrained Economic Growth; Bretton Woods; Budget Deficits;
Endogenous Money; Exchange Rates; Fiscal Policy; Full Employment; Monetary Policy;
Taxation; Tax-based Incomes Policy; Tobin Tax; Unemployment.
References
Arestis, P. and M. Sawyer (1998), Keynesian policies for the new millennium, Economic
Journal, 108 (446), 18195.
Davidson, P. (1992), Reforming the worlds money, Journal of Post Keynesian Economics, 15
(2), 15379.
Kalecki, M. (1944), Three ways to full employment, in Oxford University Institute of
Statistics, The Economics of Full Employment, Oxford: Blackwell, pp. 3958.
Keynes, J.M. (1980), The Collected Writings of John Maynard Keynes. Volume XXVII.
Activities 19401946. Shaping the Post-War World: Employment and Commodities, London:
Macmillan for the Royal Economic Society.
Lerner, A. (1943), Functional finance and the Federal debt, Social Research, 10 (1), 3851
(reprinted in W. Mueller (ed.), Readings in Macroeconomics, New York: Holt, Rinehart &
Winston, pp. 35360; citations refer to the reprint).
McCombie, J.S.L. and A.P. Thirlwall (1997), The dynamic Harrod foreign trade multiplier
and the demand-oriented approach to economic growth: an evaluation, International
Review of Applied Economics, 11 (1), 526.
Wray, L.R. (1998), Understanding Modern Money: The Key to Full Employment and Price
Stability, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Eective Demand
Eective demand is a concept of central importance in Post Keynesian economics. The principle of eective demand states that the level of economic
activity that is, aggregate nominal income and employment is determined by the level of eective demand, which is, in turn, determined by the
106
Eective demand
conjunction of aggregate supply and demand conditions. Aggregate

demand conditions are held to play a leading role in this relationship.
The determination of eective demand, and hence the levels of nominal
income and employment, results from the interaction of aggregate supply
and demand functions that describe relationships between levels of
employment and levels of business proceeds. Business proceeds constitute
receipts from the sale of final goods and services net of an allowance for
the use of capital equipment and thus represent, in the aggregate, the total
nominal income of the community. The aggregate supply function
describes a relationship between expected proceeds and the levels of
employment that firms are willing to oer on the basis of these expectations. The aggregate supply function, then, suggests that firms base their
employment oers on the value of goods and services that they think they
can sell. Not surprisingly, as expected proceeds rise, so, too, do employment
oers (see schedule Z, Figure 6). The aggregate supply function so
described can be formulated on the basis of either Marshallian or
Kaleckian microfoundations. In the first case, firms equate short-run price
expectations with the marginal costs of production in order to determine
the profit-maximizing level of output, on which employment oers are then
based. In the second case, firms set prices as a mark-up over average costs,
with both expected proceeds and consequent employment oers then determined by the volume of output that firms expect to sell at these prices
(Asimakopulos 1991, pp. 537).
The aggregate demand function, meanwhile, describes a relationship
between received proceeds and the levels of employment that generate these
receipts. It therefore relates aggregate expenditures in the economy to the
level of employment, which is a determinant of household income.
According to the aggregate demand function, higher levels of employment
are associated with higher household income, and hence higher aggregate
expenditures and business proceeds (see schedule D, Figure 6). As this
description suggests, some of the components of aggregate demand are
endogenous that is, influenced by the levels of income and employment
that they, in turn, help to determine. However, other components of aggregate demand are autonomous that is, determined independently of the
levels of income and employment.
Two important properties of the aggregate demand function mean that
it will not, in general, be identical to the aggregate supply function. The first
concerns the behaviour of endogenous components of aggregate demand.
Although the principle of eective demand associates an increase in
employment with both an increase in output and an increase in expenditures, the additional expenditures associated with any increase in employment are understood to constitute only a fraction of the value of the
Eective demand 107

Proceeds
(Y)
Y*
Z
D
Employment
(N)
N*
w*
w
Sn
Key
Daggregate demand function
Zaggregate supply function
Epoint of eective demand
Ybusiness proceeds (aggregate income)
Nemployment
wreal wage
Uinvoluntary unemployment
Sn labour supply
Figure 6
The principle of eective demand and labour market outcomes
additional output (and hence income) this increase in employment generates. This is due to Keyness fundamental psychological law, according to
which the marginal propensity to consume out of additional income is less
than one (Keynes 1936, p. 96). Hence the slope of the aggregate demand
function diers from (specifically, is smaller than) that of the aggregate
supply function, as illustrated in Figure 6.
The second important property of the aggregate demand function concerns the behaviour of autonomous components of aggregate demand.
According to the principle of eective demand, autonomous expenditures
are genuinely independent of endogenous expenditures, so that there is no
automatic tendency for variations in the latter to be oset by changes in the
former. This can be understood most simply in the context of a closed
economy with no active government sector, in which case it reduces to the
108
Eective demand
proposition that saving does not create investment. This is so for two
reasons. First, in a money-using economy, saving does not in and of itself
constitute a current demand for goods, as it does in a barter economy. In
other words, saving is not identical to investment. (Indeed, saving does not
even constitute a demand for consumption goods in some specific future
period. In an environment of uncertainty, a decision to avoid commitment
to goods and services in the present that is, to save represents an indefinite postponement of expenditure.)
Second, saving considered now as distinct from investment has no
direct impact on the rate of interest, to which investment is, in principle,
sensitive. Hence saving cannot automatically create osetting investment
expenditures through variations in the interest rate. An important feature
of the principle of eective demand, then, is its treatment of the interest
rate as a monetary variable, the determination of which is relatively autonomous from the income determination process. This autonomy is achieved
in Keynes (1936) by the liquidity preference theory of the interest rate, and
in contemporary Post Keynesian theory by central bank determination of
short-term interest rates in an endogenous money environment. Note that
the interest rate need not be absolutely autonomous from the income generation process. It is quite possible for changes in income to have some
impact on the interest rate via changes in liquidity preference or via a
central bank reaction function. Neither alters the operation of the principle
of eective demand, however. What ultimately emerges from this analysis
is a system in which the level of income (and by extension, employment)
rather than the interest rate is the key adjustment variable responsible for
equating investment and saving, according to a strict causal schema in
which changes in investment spending cause (via their impact on the level
of income) changes in saving (Amadeo 1989, pp. 12).
The level of eective demand is determined by the point of eective
demand, at which the aggregate supply and demand functions intersect
(point E in Figure 6). The proceeds resulting from this eective demand and
the associated volume of employment constitute the economys equilibrium levels of income and employment. The point of eective demand is
an equilibrium in the sense that, at this point, the expected proceeds necessary to encourage firms to oer a particular level of employment (as determined by the aggregate supply function) are exactly equal to received
proceeds at this level of employment (as determined by the aggregate
demand function). However, the importance attached to historical time
and uncertainty in Post Keynesian economics necessitate that care is taken
when interpreting the precise nature of this equilibrium relative to equilibrium constructs found in other approaches to economics (Kregel 1976).
Several salient features of the principle of eective demand as described
Eective demand 109

Proceeds
(Y)
ZD
Y*
N*
Employment
(N)
w*
Dn
Sn
Key
Daggregate demand function
Nemployment
wreal wage
Sn labour supply
Dn labour demand
Figure 7
The special case of Says Law
above are worthy of note. First, because the aggregate supply function will,
in general, coincide with the aggregate demand function at only one level
of income and employment, the principle of eective demand refutes Says
Law, according to which supply always creates its own demand. Says Law
is, in fact, revealed to be a special case, in which the aggregate demand and
supply functions are identical. In this special case, the resulting indeterminacy in the levels of income and employment is resolved by the equation of
labour supply and demand so that variations in the real wage, by establishing a market-clearing level of employment, determine aggregate income
(see Figure 7). In general, however, the aggregate supply and demand functions will coincide at only one level of employment, at which point the level
of income and the value of the real wage are also determined (the real wage
may, in fact, be constant along the aggregate supply function if the latter is
based on Kaleckian microfoundations, but will systematically decline as the
110
Eective demand
level of employment rises if the aggregate supply function is based on

Marshallian microfoundations). Whether or not the level of employment
and real wage so established corresponds to a point on the economys
labour supply schedule is an open question; in general it will not, and the
resulting deficient demand for labour will give rise to involuntary unemployment (see Figure 6). In the general case, then, variations in eective
demand cause changes in income, employment and the real wage (to the
extent that this is non-constant), and there is no automatic tendency for the
labour market to clear.
Implicit in the contrast above is the observation that, according to the
principle of eective demand, the volumes of employment and production
depend on firms anticipations of the value of output that they can sell,
rather than on the equation of the marginal physical product and marginal
disutility of labour. This draws attention to a number of other salient features of the principle of eective demand: the importance that it attaches
to the goods market rather than the labour market as the proximate determinant of the scale of economic activity; the central role it ascribes to firms
in actively setting employment and output, rather than passively responding to labour market outcomes; and the epistemological significance of
explaining economic activity in terms of expected future sales revenues
rather than the known productive capacities of factors of production.
Several important controversies surround the principle of eective
demand. In the first place, and partly as a result of Keyness original exposition, the terms aggregate demand and eective demand are frequently
confused, and there is a long-standing debate as to whether the aggregate
demand function describes the proceeds actually received by firms at dierent levels of employment, or the proceeds they expect to receive at these
various levels of employment. As regards the first of these issues, it should
be clear from the foregoing discussion that aggregate demand is a schedule
(which describes the relationship between levels of employment and received
proceeds), whereas eective demand is a point (specifically, the point where
the aggregate demand and supply functions coincide) (see also Chick 1983,
pp. 645). As regards the second issue, the aggregate demand function can
be thought of as describing either actual or expected magnitudes. Indeed, it
is useful to think of two dierent aggregate demand functions, one describing actual and the other expected proceeds (Amadeo 1989) although particular microfoundations are required in order to generate an expected
aggregate demand function that is upward-sloping (see for example,
Asimakopulos, 1991, pp. 434). Where the expected aggregate demand function intersects the aggregate supply function determines the level of employment within any production period; where the actual aggregate demand
function intersects the aggregate supply function determines the equilib-
Eective demand 111

rium position described earlier, at which expected sales proceeds equal proceeds actually received. This equilibrium will actually be achieved when the
expected and actual aggregate demand functions and the aggregate supply
function all coincide at a single level of employment (see Figure 8).
Proceeds
(Y)
Z
De
Y*
N*
Employment
(N)
Key
Dactual aggregate demand function
Epoint of eective demand
Nemployment
De expected aggregate demand funtion
Figure 8 The interaction of expected and actual aggregate demand, and

aggregate supply
More recently, Pasinetti has argued that standard expositions of the principle of eective demand only succeed in identifying the concept of a point
of eective demand, rather than anything meriting description as a principle. Pasinetti (1997, pp. 98100) identifies the principle of eective demand
with the proposition that changes in demand result in changes in output (at
least until full capacity is reached), and argues that this principle operates
at a deeper or more fundamental level than is suggested by the behavioural
relations commonly used to describe a point of eective demand (Pasinetti
1997, p. 100; 2001, pp. 3869). That the principle of eective demand is
112
Employment
associated with the proposition that macroeconomic activity is demand

determined is not controversial. But Pasinettis suggestion that this principle can be articulated at a more fundamental level of analysis than that
associated with specific behavioural relations and a particular institutional
context is contentious. It raises the issue of methodological divisions within
Post Keynesian economics, between those who favour a long-period method
of analysis in which certain core relationships obtain independently of
short-term events (in the determination of which institutions do play a
part), and those who argue that there are no long-period positions defined
and reached independently of the sequence of (behaviourally and institutionally specific) short-run outcomes leading up to them. It is not surprising, then, to find critics of Pasinettis position reasserting the fundamentally
behavioural and institutionally specific nature of the principle of eective
demand. Davidson (2001, p. 393), for example, defines the principle of eective demand as the proposition that the behavioural determinants of the
aggregate demand and supply functions dier in the institutionally specific
context of a money-using economy.
M S
See also:
Employment; Expectations; Investment; Keyness General Theory; Liquidity Preference; Rate
of Interest; Saving; Says Law; Sraan Economics; Unemployment.
References
Amadeo, E. (1989), Keyness Principle of Eective Demand, Aldershot: Edward Elgar.
Asimakopulos, A. (1991), Keyness General Theory and Accumulation, Cambridge: Cambridge
University Press.
Chick, V. (1983), Macroeconomics After Keynes, Cambridge, MA: MIT Press.
Davidson, P. (2001), The principle of eective demand: another view, Journal of Post
Macmillan.
Kregel, J. (1976), Economic methodology in the face of uncertainty: the modelling methods
of Keynes and the Post Keynesians, Economic Journal, 86 (342), 20925.
Pasinetti, L. (1997), The principle of eective demand, in G.C. Harcourt and P.A. Riach
(eds), A Second Edition of The General Theory, Volume 1, London: Routledge, pp. 93104.
Pasinetti, L. (2001), The principle of eective demand and its relevance in the long run,
Journal of Post Keynesian Economics, 23 (3), 38390.
Employment
Employment is determined by the interaction of demand and supply within
the labour market. The Keynesian and Post Keynesian emphasis has by and
large been on variations in labour demand as an explanation for variations
Employment 113
in employment and more particularly unemployment. Even within the neoclassical model, given an upward-sloping labour supply curve rather than
a vertical one a shift to the left in the labour demand function will cause
a reduction in employment. But it will not necessarily cause an increase in
unemployment in excess of the natural rate. In order to explain why there
is such a linkage, Post Keynesians appeal to inertia in the wage rate which
prevents labour market clearing. This is where much of the research in
labour economics has been. Yet it is clear from this brief synopsis that with
respect to employment per se there are at least two other issues of concern:
the magnitude of the slope of the labour supply curve and variations in the
labour supply function itself.
Explanations of the failure of the labour market to clear centre around
two distinct concepts which dier in the time frame of their impact. First,
there are those theories which relate primarily to the stickiness of nominal
wages. These theories, such as implicit contracts and menu costs, explain
why the labour market is slow to adjust to demand and supply shocks, thus
generating short-term unemployment. Second, there are those theories
such as insideroutsider, eciency wages and hysteresis, which predict
long-term deviations in the real wage from the market-clearing level. We
examine these in turn.
Changing prices, including wages, in response to every change in economic conditions is a costly business and firms find it optimal to limit this
process. With wages this is reflected with an annual pay award, which is
common in Britain, while in the US a three-year pay agreement is not
uncommon. In this case the firm will respond to changes in demand by
changing output and employment rather than wages. The work on menu
costs is associated with Gregory Mankiw (1985). George Akerlof and Janet
Yellen, Michael Parkin, Julio Rotenberg, and Olivier Blanchard and
Nobuhiro Kiyotaki have also made significant contributions.
Implicit contract theory involves long-term relationships between firms
and workers which it is not optimal to break for short-term considerations.
If workers know that they will be employed by the firm for a prolonged
period then they are prepared to accept a fixed wage lower than their marginal productivity in return for this guaranteed stability. The reason why
workers and firms might prefer implicit contracts with relatively fixed
wages is partly a transfer of risk from the more risk-averse individual to the
less risk-averse firm. The firm is acting as an insurer against fluctuations in
their income attributable to the business cycle. A seminal paper is by Baily
(1974). Other important contributions include Costas Azariadis and
Edmund Phelps.
A slightly dierent concept is that of eciency wages (Solow 1979; see
also Carl Shapiro, Akerlof and Yellen). The central assumption is that there
114
Employment
is a benefit as well as a cost to a firm of paying a higher wage. There are

several reasons why this might be the case. First, a higher wage can lead to
a healthier, better-nourished and therefore more productive workforce.
Second, a higher wage can increase loyalty among the workforce and hence
induce greater eort. Third, higher wages can help in monitoring workers
eort in situations where this is imperfectly observable. Where the wage is
equal to the market-clearing wage and where there are relatively abundant
job opportunities, workers are indierent to the possibility of losing their
jobs and hence might be tempted to shirk. Paying higher wages will make
workers keener to retain their well-paid jobs. A variation on this concept is
linked to the diculties associated with the hiring of workers. Given that
not all workers are of equal ability, a firm wishing to hire high-quality
labour may find itself frequently disappointed and often faced with the cost
of dismissing unsatisfactory workers and rehiring others. To minimize such
costs the firm may pay above the market wage, reasoning that workers know
their own abilities and poor workers will recognize that this is a job that
they will soon be fired from and hence not apply. This is therefore an equilibrium concept; it does not explain sticky wages per se, it explains why
wages may be set above workers marginal productivity due to informational asymmetry. It does, however, indicate that employment will be lower
than would be the case without this form of market imperfection.
Hysteresis as applied to unemployment is often traced to the influential
paper by Blanchard and Summers (1986), although in reality the term hysteresis can be traced all the way back to Joseph Schumpeter at least.
Hysteresis challenges the natural rate assumption of neoclassical economics in arguing that even in the long run there are factors which may prevent
unemployment returning to its equilibrium level following a (deep) recession. First, a prolonged spell of unemployment might see a deterioration of
the workers skills, thus making re-entry into the labour force more dicult.
Second, it may induce a change in attitudes to work and reduce the incentive to find employment. Another possible explanation for observed hysteresis is entry into the hidden economy as the formal economy declines,
with entrants reluctant subsequently to exit when given the chance. Thus
eectively in economics hysteresis is a supply-side concept; it indicates that
following a recession the labour supply curve will shift to the left and will
shift back only slowly, if at all. Technically the implication of this is that
employment is a random walk with drift, that is, the change in employment
is equal to a constant term plus a stochastic, white noise error term. Post
Keynesians are sometimes categorized as being non-technical. This is
partly a consequence of their belief that theory must be realistic, and mathematical models inevitably involve simplification. Yet the econometric work
aimed at determining whether a series is hysteretic, which is possibly linked
Employment 115
to its stationarity properties, has made sophisticated use of recent advances
in econometrics. This is also evident in other areas of analysis too.
The final approach we shall consider is that related to insidersoutsiders
and trade union bargaining (McDonald and Solow 1981). This centres
around the concept that those who are already employed (insiders) have
more say in wage bargaining than those out of work (outsiders). Often the
former are represented by a trade union, which bargains mainly with their
interests in mind. The fact that lower wages would help the outsiders does
not figure prominently in this decision. This leads to some predictable and
testable outcomes, such as the greatest upward impact on wages being in
countries with strong trade unions and decentralized bargaining.
In 1960, employment in the US totalled 74 million people. By 2000 this
had risen to nearly 140 million. The bulk of this increase came from an
increase in the working population aged between 15 and 64 which
increased from 108.4 million to almost 180 million. However, another significant factor was the increase in participation rates, particularly female
participation rates. But the market dynamics are complex, and there is a
considerable literature which links the decline in the wages of low-skilled
men in the 1980s with this rise in female participation which should generate a reduction in male labour supply. What is beyond dispute is that this
rise in female participation rates has been oset to some extent by an even
sharper decline in the participation rate of older men, both through a trend
to early retirement and a reduction in men working after the retirement age
of 65. The impact of these latter changes is mostly to have shifted the
labour supply function inwards.
The second point we turn to is the possibility of hysteresis in the demand
curve for labour. By this is meant that the relation of this demand function
to aggregate demand for goods and services may shift following a recession.
In other words it is time dependent. This argument is linked to changes in
the level of capital following a recession, through either capital scrapping
or the reduction in investment, even in some cases replacement investment.
Hence following the recession the aggregate supply curve shifts to the left,
and an increase in the aggregate demand for goods will be met by increased
inflation and imports rather than increased output and employment
(Hudson 1999). In terms of the labour market, the demand function for
labour itself becomes sticky. In this respect it has been argued that US
bankruptcy laws, which place much more emphasis on the retention of
capital and firms as going concerns, are better for long-term employment
prospects than European bankruptcy laws.
In approaching this survey I have, in common with most of the textbooks, extended the analysis to include all theorists who reject in part the
neoclassical paradigm as applied to the labour market, even though they
116
Employment
might not describe themselves as Post Keynesians. This leads us into including in the ambit of this entry theories which Keynes himself would have
clearly rejected as irrelevant, and in some cases actually did so. Thus the
insideroutsider theory is arguably simply a new bottle for the old wine of
labour market clearing being frustrated by union power a wine which was
flowing steadily in the 1920s and 1930s and which Keynes had diculty
swallowing.
Philip Arestis suggests that there are three traditions to Post Keynesian
analysis: (i) the importance of eective demand failure, (ii) the role of
uncertainty and (iii) the role of institutions, particularly in dealing with the
problem of bounded rationality. The principle of eective demand, the
concept that demand constraints are the dominant factor in explaining
variations in output, is the backbone of the Post Keynesian approach. But
to understand why this leads to unemployment rather than simply to a
decline in employment we need to appeal to the other two concepts.
Arestis has also argued that social classes are essential to the second of
these approaches, which is essentially Marxian, in adapting Marxs reproduction system to solve the realization problem, and that the third
approach is also based upon the dynamic and power/class struggle of economic systems, in which income distribution plays a key role. This may well
be so, yet equally concepts such as eciency wages, implicit contracts, even
hysteresis have merits in their own right which appeal and inform beyond
Post Keynesians.
It is also noteworthy, perhaps, that in areas where the spirit of the traditions of Karl Marx, Thorstein Veblen and Michal- Kalecki might be
thought to have particular relevance, for example, labour force participation and the shadow economy, Post Keynesians have largely been absent.
Thus, for example, with respect to household production functions the neoclassical school has been given virtually a free run. This is a reflection of the
fact that to a considerable extent the Post Keynesian analysis of employment is in fact an analysis of unemployment. This is unfortunate: knowledge is forged in the heat of debate. Post Keynesians should have the
confidence and the curiosity to expand their horizons.
J H
See also:
Eective Demand; Full Employment; Income Distribution; New Keynesian Economics;
Wages and Labour Markets.
References
Baily, M.N. (1974), Wages and employment under uncertain demand, Review of Economic
Studies, 41 (1), 750.
Endogenous money 117

Blanchard, O.J. and L.H. Summers (1986), Hysteresis and the European unemployment
problem, NBER Macroeconomics Annual, 1, 1578.
Hudson, J. (1999), A generalized theory of output determination, Journal of Post Keynesian
Economics, 21 (4), 66378.
Mankiw, N.G. (1985), Small menu costs and large business cycles: a macroeconomic model
of monopoly, Quarterly Journal of Economics, 100 (2), 52937.
McDonald, I.M. and R.M. Solow (1981), Wage bargaining and employment, American
Economic Review, 71 (5), 896908.
Solow, R.M. (1979), Another possible source of wage stickiness, Journal of Macroeconomics,
1 (1), 7982.
Endogenous Money
Money may be defined as the asset generally accepted as the means of
payment and medium of exchange. The particular asset that is conventionally bestowed with general acceptability has changed enormously over
time. For conceptual clarity concerning supply characteristics, it is necessary to distinguish commodity, fiat and credit money.
A wide variety of commodities have been used as money, items as diverse
as shells, beads, rice, salt and cattle. The general acceptability in exchange
of commodities such as gold and silver evolved gradually over prehistory.
Both possessed properties attractive for a money asset, homogeneity, durability, divisibility, a high value to weight ratio, and reproducibility under
sharply increasing costs, so their short-run supply was broadly given.
Money does not grow on trees. Precious metals are assets to their holders
and liabilities to no-one. Commodity money is a physical asset and not a
financial claim.
Fiat money is a financial asset, the non-interest-bearing debt of the
government. It derives its value because it has been declared legal tender in
settlement of debts and taxes, by government fiat. Once the general acceptability of precious metals was established, their main drawback was diculty in verifying their value. The development of fiat money substituted a
dicult to counterfeit certification of value for a full-bodied metal coin.
Since their creation lies completely outside the lending and borrowing
process, the quantity of commodity or fiat money denotes nothing about
the outstanding volume of credit. As a result the supply of commodity and
fiat money can be assumed exogenous, and independent of changes in the
demand for money. It appeared reasonable to assume that, in response to a
change in the supply of money, prices and incomes would adjust until the
exogenous stock was again willingly held. This was the original basis for the
Quantity Theory of Money, which maintained that changes in the quantity of money were responsible for changes in the level of prices and money
income.
118
Endogenous money
Banking originated when Italian goldsmiths first took the strictly illegal
step of lending out someone elses gold. The crucial banking innovation
was the making of bank IOUs payable to the bearer rather than to a
named individual. This led to the discovery that, so long as public confidence in the liquidity of deposits was maintained, a banking house of sucient repute could dispense with the issue of coin for most transactions, and
instead issue its own bank notes.
Credit money is the liability of the issuing bank, and is backed by borrowers liabilities (IOUs) in the banks possession. The supply of credit
money varies with changes in the demand for bank credit. Deposits are
continuously created and destroyed through the granting and repayment of
loans. Credit money is never in excess supply, providing the public remains
confident that all deposits will be redeemed in fiat money on demand.
Unfortunately the process by which credit money is supplied is somewhat
non-transparent. The banking system operates under strict rules imposed by
the central bank. In most economies the central bank requires commercial
banks to maintain a minimum required ratio of reserves to deposits. This
must be complied with. It thus appears that the supply of credit money is
controlled by the total volume of bank reserves supplied by the central bank.
Mainstream theory postulates that the total supply of bank deposits is
governed by the total quantity of reserves made available by the central
bank. The central banks total liabilities (outstanding notes and coin and
bank reserves held with the central bank) are termed the high-powered
base. Since the base bears a constant ratio to the stock of money (m
M/B) it is believed to support a multiple quantity of deposits. The change
in bank loans (L) and in the money supply (M) may be written as a
multiple m of the change in the high-powered base (B). Reserves cause
deposits: mB LM.
But this equation does not explain how credit money is supplied. It is
simply a rewriting of the identity (mM/B). Identities say nothing
about the direction of causality. Post Keynesians maintain that causality
goes from changes in bank loans (L) to changes in the money supply (M)
to changes in the base (B): LMB (Loans cause deposits).
The mainstream view that central banks control the supply of money
hides the fact that banks are essentially retailers of credit, not portfolio
managers. Banks sell credit, and create deposits in the act of making loans.
Like other retailers, the quantity of credit they sell depends on demand.
Post Keynesians substantiate their theory of reverse causality, that the
direction of causality goes from changes in loans to changes in the base, so
that the supply of credit money is endogenous and credit driven, as follows.
(For alternative but similar Post Keynesian explanations, see Lavoie 1992;
Moore 1988, 2003; Rochon 2001; and Wray 1990.)

Credit money evolved from commodity and fiat money in parallel with
the development of capitalism and the private ownership of the means of
production. It was no accident that commercial banking first evolved in
Renaissance Italy, in response to the demand for working capital by early
Italian merchant traders. Business demand for working capital was the
major force behind the development of commercial banking.
Production takes time. In market economies production costs must be
incurred and paid before the receipt of sales proceeds. Business costs of
production constitute business demand for working capital, for which firms
must somehow find financing. The owners of firms bear all the risk of production under uncertainty. They must pay out money to purchase productive factors before they receive profits from expected future sales.
Credit is demanded by firms to finance increases in the demand for
working capital during the production-sales period. Increases in the
volume of output require proportional increases in working capital to
finance the higher value of inventories of goods in process.
The acceptability of bank deposits as a payments instrument depends on
the publics confidence that deposits are always exchangeable for legal
tender on demand. Banks volitionally demand cash reserves to maintain
the general acceptability of deposits. Required reserve ratios above the ratio
of reserves to deposits that banks desire to maintain on liquidity grounds
act as a tax on banking, impairing the eciency of bank intermediation
and their international competitiveness.
Banks are price setters and quantity takers in their lending and deposit
markets. So long as they remain within their assigned overdraft ceiling, the
amount of loans taken out is the decision of the borrower, not the bank.
Firms borrow primarily to finance their demand for working capital, of
which the largest component is the wage bill. Loans are repaid after firms
receive their sales proceeds, the notion of the monetary circuit (see
Deleplace and Nell 1996 and Graziani 1989). An increase in the wage bill
has a greater than unitary eect on total bank lending. Banks lend to
households in addition to firms, and household borrowing is positively
related to wage income.
The supply of credit money is endogenously credit driven. Banks only
lend to borrowers whom they believe can repay their debts. Bank lending
ocers make a discretionary judgement of the maximum amount an individual borrower will be able to repay, based on their estimates of the borrowers three Cs: collateral (liquid wealth), credit (pecuniary income) and
character (borrowing history). They provide overdraft facilities (credit
lines) up to this ceiling or limit.
When a banker makes a loan he or she credits the borrowers account.
The supply of credit money responds to changes in the demand for bank
120
Endogenous money
credit. Bank loans operate like credit cards. Modern banks provide creditworthy borrowers with substantial overdraft facilities, and then make credit
available on demand up to their credit limit. (The credit-utilization rate in
developed economies fluctuates around 50 per cent; see Moore 1988,
chapter 2.) Most poor households and small firms, and in developing economies the entire informal sector, are credit constrained. They form
Keyness perpetual fringe of unsatisfied borrowers. Banks regard lending
to them as too risky, so they receive no credit. Quantity rationing is inherent in credit markets.
For the banking system as a whole, total loans equal total deposits. But
individual banks must match their supply of funds to their loan demand.
Since the innovation of certificates of deposit (CDs), deficit banks are able
to borrow funds directly from surplus banks in wholesale markets, reducing their demand to hold liquid reserves.
The conclusion that the money supply is credit driven is strongly supported empirically for many countries over dierent periods (see Moore
1988, chapter 9).
Mainstream theory regards the level of interest rates as determined by
the real forces of supply and demand for loanable funds, and by wealthholders liquidity preferences. But in all modern economies the level of
short-term interest rates is exogenously set by the central bank.
Cash reserves are supplied to the banking system as demanded by the
central bank, in its role of residual supplier of system liquidity. It sets the
interest rate at which it lends reserves to the banking system to realize its
stabilization objectives. Central banks have great discretion over the level
of interest rates they set, even in highly open economies. Only when the
central bank adopts a fixed exchange rate regime does it lose its ability to
set the short-term rate; the domestic rate then becomes equal to the foreign
rate, set by the central bank of the largest foreign economy.
Central bank open-market operations provide the banking system with
slightly less reserves than they are required to hold. In this manner it keeps
banks in the Bank, and forces them to borrow reserves at the margin from
the central bank. This enables the central bank to ensure that the overnight
rate it charges the banking system for the loan of reserves, termed bank
rate in most countries, becomes the short-term market interest rate (in the
US it is termed the federal funds rate).
Banks administer their lending and deposit rates, and attempt to meet all
loan demand that results, so long as borrowers remain within their assigned
overdraft ceiling. Lending rates are set as a mark-up over the central banks
lending rate. The size of the mark-up depends on their market power in
dierent markets, reflecting their estimate of the interest elasticity of the
demand for credit. Administered lending and deposit rates are changed

infrequently, and price leadership is common. Over the period between
changes in bank rate, the supply of bank credit is horizontal, at the interest rate set exogenously by the central bank.
Changes in the short-term interest rate have become the central banks
chief exogenous policy instrument to achieve its stabilization goals. Shortterm rates are varied procyclically over the business cycle, by an amount
depending on the authorities policy reaction function (see Moore 2003),
recently termed a Taylor Rule.
Thus the supply of credit money is endogenously demand determined
and credit driven. Loans create deposits, banks decide who is creditworthy,
creditworthy borrowers decide on the amounts they wish to borrow, and in
so doing determine the supply of credit money, and the central bank sets
the short-term interest rate as its policy instrument. Purchasing goods and
services with deposits transfers the ownership distribution of deposits
among economic units, but does not change the quantity. Once created,
deposits are always accepted, so long as they retain their general acceptability. Deposits are reduced only by encashment into currency, by the repayment of bank loans and by the scale of securities by the banking system.
The volume of lending by the banking system determines the volume of
lending to the banking system.
B M
See also:
Banking; Central Banks; Circuit Theory; Credit Rationing; Development Finance; Monetary
Policy; Money; Rate of Interest.
References
Deleplace, G. and E. Nell (eds) (1996), Money in Motion: The Post Keynesian and Circulation
Approaches, London: Macmillan.
Graziani, A. (1989), The Theory of the Monetary Circuit, London: Thames Papers in Political
Economy.
Lavoie, M. (1992), Foundations of Post-Keynesian Economic Analysis, Aldershot, UK: Edward
Elgar.
Moore, B.J. (1988), Horizontalists and Verticalists: The Macro-economics of Credit Money,
New York: Cambridge University Press.
Moore, B.J. (2003), Shaking the Invisible Hand: Complexity, Endogenous Money and
Exogenous Interest Rates, Cheltenham, UK, and Northampton, MA, USA: Edward Elgar.
Rochon, L.-P. (2001), Credit, Interest Rates and the Open Economy: Essays on Horizontalism,
Cheltenham, UK, and Northampton, MA, USA: Edward Elgar.
Taylor, J. (1999), Monetary Policy Rules, Chicago: University of Chicago Press.
Wray, L.R. (1990), Money and Credit in Capitalist Economies: The Endogenous Money
Approach, Aldershot, UK: Edward Elgar.
122
Environmental economics
Environmental Economics
Over the past twenty years or so environmental economics from being a
fringe activity has become one of the most active areas of economic
research. (In this entry the term environmental economics is used broadly,
to include the closely related area of natural resource economics.) It has
become a major, even dominating, influence within significant areas of
policy debate, including momentous global issues such as climate change
and biodiversity loss. Mainstream environmental economics is currently
dominated by the neoclassical paradigm in the ways in which it formulates
and analyses the two key areas with which it is concerned: the valuation of
environmental assets and the design of policy instruments to manage those
assets. These are bought together in the study of sustainable development:
how is it to be defined and achieved, if, indeed it is desirable? Thus environmental economics is essentially a branch of applied welfare economics. In
some respects, environmental economics represents a rather extreme interpretation of the neoclassical paradigm, with its belief in the possibility of
extending, with reasonable reliability, individual valuations to all sorts of
non-marketed commodities, with its definition of environmental problems as essentially flowing from market failures, and with its advocacy of
the ecacy and desirability of incentive-based policy instruments to
correct for these failures.
Partly due to its somewhat extravagant faith in the neoclassical paradigm
and partly, also, because of the necessary interface between environmental
economics and the natural sciences, mainstream environmental economics
has not been without its critics. Some of this criticism is simply misplaced
(for example, that environmental economics cannot properly account for
the life cycle of products), and easily rebutted by any well-trained neoclassical economist. But some is fundamental. This is especially true of those
criticisms which challenge the foundations of neoclassical approaches to
the environment, and which thence seek alternative accounts of sustainability based on physical or natural processes intrinsic to the environment,
such as energy usage (Georgescu-Roegen 1971) or biological resilience
(Common and Perrings 1992). Some of these accounts aspire to create an
entirely new form of economics based, for example, on a redefinition of the
concept of scarcity or value.
Leaving this aside, for the moment, what is conspicuously lacking in the
debate is a serious attempt to draw ideas from heterodox schools of thought
within economics, broadly defined, in order to criticize and reformulate
environmental economics. The only significant exceptions are as follows.
First, some moves have been made to extend neo-Ricardian models of production and growth to incorporate some process-related natural resource
123
and environmental components (Kurz and Salvadori 1995, chapter 12).

Second, richer, psychologically-based or socially-embedded accounts of
human behaviour have been drawn on to criticize the appropriateness of
attempts to extend neoclassical valuation processes over non-market
domains (Kahneman and Knetsch 1992; Sago 1988). However, beyond
these, it seems obvious that Post Keynesian economics can provide insights
which have the potential to provide a powerful critique of neoclassical environmental economics and pointers to a reformulation of the subject. This
entry attempts to outline a few of these insights. It uses the term Post
Keynesian economics in a comprehensive sense, so as to include and
develop both the steps just mentioned, along with others. All this is tentative and very much represents work in progress.
In the neoclassical view, environmental problems are just one species of
externality and are to be costed at the price which an ecient market
would impute to them: they would not exist if markets were complete and
in equilibrium. This seems to many to fail, in some sense, to grasp the real
existence of environmental problems independent of their specification in
an economic model. It might be a reason for adopting one of the alternative accounts, mentioned above, of what might be called environmentallyembedded sustainability in order to define the nature of environmental
problems, with all the foundational issues thus entailed. The present entry
proceeds more pragmatically, and attempts in part to formulate a debate
between neoclassical and possible Post Keynesian perspectives on the
environment.
Does the Post Keynesian approach encompass the concept of externalities? Presumably not. To use two (or three) arguments, which may or not
be consistent with each other: what sense would the neoclassical notion
of allocative eciency make in, first, a (non-trivial) monetary economy
where markets were necessarily incomplete or in continuous disequilibrium, or, second, in an economy where prices were (re)production prices,
Sraan or Kaleckian, and not indices of scarcity? To this it might be
objected that neoclassical general equilibrium is an ideal-type construct
which specifies the necessary conditions for allocative eciency, and this
normative status is untouched by Post Keynesian arguments. This raises
deep questions about the nature of economic models, which are not
pursued further here.
A Post Keynesian argument which appears more secure against this sort
of objection might run as follows. The Post Keynesian perspective on the
nature of prices would apply to the attempt to use supposedly allocatively
ecient prices to value environmental assets and damage to them. In many,
perhaps most, of the cases which are of most interest to environmental
economists, there are no observable or even imputable prices of any sort to
124
use in such valuations. One widely used procedure in such cases is to use the
so-called contingent valuation method, eliciting prices by questioning
people about their willingness to pay for environmental benefits or to
accept losses. As noted above, this has provoked a variety of criticisms,
partly because of what appear to be irreducible anomalies, but a Post
Keynesian one might run along these lines. The answers given in contingent
valuation surveys could represent an attempt by respondents to formulate
a response based on the prices that people know in their everyday economic
lives. So what if these prices are not, for one or other of the reasons given
above, to be interpreted as meaningful indicators of underlying preferences
and relative scarcities? We might ask: where does a persons notion of an
appropriate price come from if not from social practice?
It is indisputable that many environmental problems, however defined,
involve extended time horizons and extreme uncertainty, and that these are
closely related. This is fertile territory for Post Keynesians. In mainstream
environmental economics, time is routinely dealt with by discounting. This
is a source of much criticism from environmentalists, for familiar practical
and ethical reasons (Broome 1992). For Post Keynesians, similar arguments might be made as in the case of prices in general. In what sense can
an ecient interest rate be defined bearing in mind, say, the capital controversies or the concept of the interest rate as an essentially monetary phenomenon? Here, an attempted neoclassical rebuttal would not be so
convincing: the notion of an interest rate which somehow encapsulates
intertemporal eciency would be regarded by many Post Keynesians as
simply meaningless. Needless to say, this is contentious.
In dealing with risk, standard environmental economics generally
assumes a world of calculable probabilities. Post Keynesians would, of
course, reject this in favour of radical uncertainty, which undoubtedly characterizes many environmental problems. Not only does this undermine the
specifics of much environmental modelling, forecasting and management,
but it also links with the foregoing questions to do with the nature of interest rates and prices so as to lend additional support to a more comprehensive critique of neoclassical environmental economics.
More positively, there is a need for a proper integration of natural
resources and environmental assets into a well-formulated model of a monetary economy. Here Post Keynesians have a real chance to develop an
innovative approach, perhaps building on an own-rates analysis.
The standard neoclassical model of ecient resource extraction relies on
asset valuations based on arbitrage across asset returns. Financial instruments are introduced as simple comparator assets. A Post Keynesian would
regard this as a quite inappropriate way to capture the essential characteristics of a monetary economy.
125
Sustainability is a highly debatable concept. Its status within neoclassical environmental economics is not entirely clear: it is essentially a side condition, rather than intrinsic to the logic of the model. At all events, the core
of the concept is that some measure of welfare is bounded from below over
time. This is often expressed in terms of maintaining an appropriate aggregate capital stock. Welfare is ultimately dependent on the return to this
stock. The capital stock is very broadly defined, to include natural resource
and environmental assets, alongside physical, human and even social
capital. It should be noted that this framework is very widely used, even by
those who are dismissive of neoclassical environmental economics. Indeed,
one of the more common mistaken criticisms of environmental economics
is that it does not use a comprehensive enough definition of capital. (This
needs to be distinguished from the criticism that the market-based values
used in aggregation are inappropriate.) A Post Keynesian would argue,
instead, that such aggregation procedures are inherently flawed. There is a
need, however, to explicitly extend the so-called Cambridge critique to
encompass natural resources and environmental capitals. In addition, the
problems, already discussed, with assigning allocatively ecient prices to
the components of such capital stocks, and uncertainties in future stocks,
would be further ingredients in a comprehensive critique.
Many of the mainstream accounts of sustainability-as-maintainingaggregate-capital strengthen the criterion by requiring some individual
components of the aggregate to be maintained as well, on the grounds that
the weaker criterion overestimates the possibilities of substitution within
the economy, though others are more sanguine. A Post Keynesian would
presumably have no problem with models of production that assume
limited substitutability. But introducing it as an assumption does raise
questions about the coherence of the neoclassical model of sustainability,
which do not seem to be very clearly appreciated.
In understanding the possibilities for long-run sustainability, an area
which needs further exploration is the integration of natural resource and
environmental assets into growth models, especially those with endogenous
innovation (Aghion and Howitt 1998, chapter 5). Some of the newer work
on endogenous innovation and growth has strong similarities, in some
respects, with Kaldorian models, but otherwise relies on questionable neoclassical modelling of representative agents. One particular aspect that has
barely been investigated, which again has a strong Kaldorian flavour, is the
relationship between environmental performance and the sectoral and
spatial structure of the economy. What are the relative natural resource and
pollution intensities of production and consumption processes in the
primary sector, in manufacturing and in services, and at various population
densities? Both innovation and structure have acquired crucial significance
126
Equilibrium and non-equilibrium
in evaluating the sustainability of the new information economy, and Post

Keynesian, and especially Kaldorian, approaches would seem to have
much potential here.
A W
See also:
Capital Theory; Kaldorian Economics; Production; Rate of Interest; Time in Economic
Theory; Uncertainty.
References
Aghion, P. and P. Howitt (1998), Endogenous Growth Theory, Cambridge, MA: MIT Press.
Broome, J. (1992), Counting the Cost of Global Warming, Cambridge, UK: White Horse Press.
Common, M.S. and S.C. Perrings (1992), Towards an ecological economics of sustainability,
Ecological Economics, 6 (1), 731.
Georgescu-Roegen, N. (1971), The Entropy Law and the Economic Process, Cambridge: MA:
Harvard University Press.
Kahneman, D and J.L. Knetsch (1992), Valuing public goods: the purchase of moral satisfaction, Journal of Environmental Economics and Management, 22 (1), 5770.
Kurz, H.D. and N. Salvadori (1995), Theory of Production: A Long-period Analysis,
Cambridge, UK: Cambridge University Press.
Sago, M. (1988), The Economy of the Earth, Cambridge, UK: Cambridge University Press.
Equilibrium and Non-equilibrium

Three facts relating to economic analysis in general are relevant in understanding equilibrium and non-equilibrium. First, the primary purpose of
economic analysis is to explain real economic phenomena (prediction, of
course, may be a byproduct). Second, one common method by which economic explanation proceeds is through the construction of economic
models. And third, model building necessitates abstracting from reality and
concentrating on the smallest number of forces that adequately represent
what is actually happening. In this context, consider the notion of equilibrium first.
Equilibrium is a feature of a model. A collection of variables is in equilibrium if they are at rest. That is, all forces or laws in operation that might
influence the values those variables take on balance each other out and
there is no tendency for the variable values to change. The idea of equilibrium was imported from the physical sciences where it was (and is) applied
with respect to the properties of physical, nonsentient objects. In the economic context, the focus of equilibrium has been, and continues to be, on
human action derived from mental decision processes. Thus equilibrium
obtains when no decision maker, to the extent that his or her action has
been appropriately captured in the model, has even the slightest motivation
to change any plan or action.
127
Traditionally, the notion of equilibrium has been one of the most, if not
the most important organizing feature of economic analysis. In many
instances, either equilibrium in a model is thought to prevail, or it is taken
to be the end towards which everything is moving. Explanation based on
the former often interprets observations of the real economic world as equilibria. Thus, for example, each pricequantity data point arising in an isolated market may be construed as located at the intersection of the demand
and supply curves assumed to be in play at the moment it was observed,
and may therefore be explained as the outcome of the interaction of the
forces of demand and supply. Explanation founded on the latter frequently
interprets observations as lying on a time path that converges to equilibrium, as is the case when, say, cobweb models with stable equilibria are
taken as the representation of reality. From either perspective the end result
(here, the equilibrium), is independent of the movement towards it (should
that movement be relevant).
Since the presence of equilibrium requires no change in the variable
values at equilibrium, and since change, or a lack thereof, can only be discerned over time, the idea of time is fundamental to that of equilibrium.
Moreover, the nature of the concept of time employed has to be such as to
make it possible to recognize circumstances in which change in the relevant
variable values is absent. The notion of time usually invoked for this
purpose is called logical time. Logical time merely provides a way of
ordering events without reference to the actual passing of time. Although,
in this manner of representing reality, past events come before present
events and present events come before future events, the dierent possibilities and significance for the spacing of those events, along with the fact that
past, present and future events all have dierent qualities in relation to
human abilities to know and experience them, are ignored. All events (past,
present and future) are assumed to be completely knowable at least probabilistically. This means that plans can be assumed to be carried out on the
basis of correct knowledge and expectations that will, on average, turn out
to be correct. As a consequence, it is possible to envisage behavioural and
expectational variable values that reflect realized plans and hence do not
require alteration as the system represented by the model moves across
past, present and future states. Logical time, then, along with the requirement of full knowledge that goes with it, is essential to the traditional
concept of equilibrium as described above. In this form, equilibrium has
been invoked by economists independently of time (though time, as just
described, is implicit), as existing through time, as temporary or changing
as time passes, in terms of unchanging growth rates and full-employment
growth paths, and in reference to the short and long runs and to both microeconomic and macroeconomic phenomenon.
128
But in spite of its usefulness in, and widespread appearance over almost
two centuries of economic discourse, the concept of equilibrium has little
place in Post Keynesian economics. This is because Post Keynesian economics, in part, views the workings of the real economic system as a
process of continuous and organic change (Galbraith 1978, p. 8) that is so
fundamental, pervasive and dynamic that that system cannot be corralled
by any notion of equilibrium. The vehicle for the expression of this change,
in addition to a more sensitive regard for the institutional structures in relation to which the actual economy operates, is the vision of historical, as
opposed to logical, time that informs virtually all Post Keynesian analytical constructions.
By the phrase historical time is meant time that is actually experienced
by human beings, in which each moment in history is unique, in which
knowledge of past events is necessarily fragmentary and variable depending on each persons perception of them, and in which the occurrence of
future events and their properties is not only unknown even probabilistically, but is also unknowable. As time passes, individuals change not only
with respect to the knowledge in their possession or in their epistemic
status, but they also experience unforeseeable modifications in their economic endowments and in their perceptions of external institutional structures, environments and the possibilities of action taken by others. Because
planned behaviour continually changes with these variations, it is not possible for behavioural variables to remain constant through time. Hence the
concept of equilibrium, at least in its conventional connotation, is both
irrelevant and meaningless.
The organizing feature that replaces the notion of equilibrium in Post
Keynesian economics may be referred to as non-equilibrium. An analysis
is organized in reference to non-equilibrium if the notion of time upon
which it rests is historical. Because of the changing nature of the subjects
of inquiry, because of the fluidity of economic endowments and expectations that market processes generate, and because of the continually evolving character of institutional structures, such an analysis constructed for
any moment or period is dierent from that constructed for any other.
Moreover, non-equilibrium analyses permit the end result of a process (not
an equilibrium) to be influenced by the means of achieving it. They also
allow for the presence of non-probabilistic uncertainty and are capable of
including in their explanatory reach phenomena that are subject to that
uncertainty. Thus, for example, in the context of decision making that
accounts for (non-probabilistic) ignorance of the future, a richer complex
of forces can be incorporated into the analysis and description of economic phenomena. And the recognition of money as a historical time and
non-probabilistic uncertainty phenomenon, in which actual time passes
129
between the receipt and disposal of funds, permits addressing the reality
that the set of forces obtaining at the time of the funds receipt might be
dierent from those at the time of their disposal. The latter, in turn, may
give rise to changes in liquidity preference and in the level and stability of
expenditure streams, and macroeconomic problems in which labour
markets do not clear may result. Like traditional economics, explanation
from the Post Keynesian perspective may also proceed by constructing
models. But, as this discussion suggests, these models are rather dierent
in character from those of traditional economics. (For a more complete
discussion, see Katzner 1998.)
At first, in the construction of an explanation of a phenomenon for, say,
the period between moments t0 and t1, where t0 t1, Post Keynesian or
non-equilibrium analysis may be conceived of as proceeding analogously
to that which would arise under the equilibrium alternative: a model may
be built up, its solutions or time paths studied, and one time path identified
with observed reality between t0 and t1. Then, by pursuing the latter time
path beyond t1, one possible description of what could happen next may
be provided. It is important to understand, however, that although nonequilibrium-analytic models constructed in this way might seem similar to
their equilibrium-analytic counterparts, the dierence between them
remains significant and far-reaching. For unlike the models of equilibrium
analysis, a non-equilibrium-analytic model itself, that is, the variables,
parameters and relations of which it is composed, would be thought of as
time dependent. Once t0 or t1 changes, one could not, for the reasons
described above, expect the same model to be appropriate. Hence the fixed
parameters do not remain fixed, and the structural relations themselves dissolve: if t1 were to increase, then the history of reality is modified by the
passage of time and explanations of occurrences after t0 undergo such profound change (due to the unforeseen and unpredictable novelty at t1 that
enters the fabric of real life after t1) that the analytical structures of those
explanations are unlikely to hold up in its wake. Shackle (1974, p. 42) has
referred to this process in general as kaleidics. Clearly the methodology of
non-equilibrium analysis does not allow formal prediction as permitted by
the methodology of equilibrium analysis.
Still, stretches of time may unfold during which real-world newness does
not appear to impinge substantively on the particular phenomena under
investigation. The lack of impact could be reflected in at least two ways.
First, it may be that the equations of a model seem to be roughly stable over
time in the sense that for a while, as t1 expands, a single time path (stationary or otherwise) generated by the model continues to approximate observations of reality reasonably well. Here traditional equilibrium analysis,
though coming from a dierent methodological perspective while using the
130
same variables, parameters and relations, provides a formal picture of the

real world identical to that furnished by Post Keynesian or non-equilibrium
analysis. Furthermore, it is only in unusual cases like these that the possibilities, exposed by the model, of what could happen in the future actually
transpire.
Second, and perhaps even less likely, stability over time could also arise
with respect to the solution values of the variables even as the equations of
the model modify. Thus it may happen that all observed values between t0
and t1 are essentially similar. Hence reality between t0 and t1 could be
viewed as in a pseudo-stationary state: novelty occurs, but either not to a
sucient extent or not in ways that significantly aect the values of the variables in the analysis. Bausor (198283, pp. 1737) calls this historical equilibrium. In such a case, what is seen can be explained either in terms of a
traditional equilibrium analysis with a single model containing a unique
stationary state and with all of the assumptions that that entails or, in the
language of non-equilibrium analysis, with a model whose equations
might modulate but whose solutions remain essentially the same over time.
During periods of transformation, the two approaches clearly supply
quite distinct explanations of real-world phenomena. On the one hand, traditional equilibrium analysis can only explain unforeseen change by asserting, after the fact, that outside forces caused alteration in functions or
parameter values. But the idea of a continually modulating equilibrium
responding to, say, repeated parametric variation is not a very satisfying
way of conceptualizing the eects of novelty. Non-equilibrium analysis, on
the other hand, cannot provide much of an understanding of unforeseen
change either. Yet the acknowledgement of such change is part of the internal structure of its methodological tissue. Non-equilibrium analysis
expects change and leaves room for it. Its emphasis is on process and the
present state, and on what might happen subsequently. By comparison,
equilibrium analysis, with its focus on the end result towards which time
paths, if not already there, converge, makes room for change only after it
has been observed.
D W. K
See also:
Dynamics; Expectations; Non-ergodicity; Time in Economic Theory; Uncertainty; Walrasian
Economics.
References
Bausor, Randall (198283), Time and the structure of economic analysis, Journal of Post
Galbraith, John Kenneth (1978), On Post Keynesian economics, Journal of Post Keynesian
Economics, 1 (1), Fall, 811.
Exchange rates 131

Katzner, Donald W. (1998), Time, Ignorance, and Uncertainty in Economic Models, Ann
Arbor: University of Michigan Press.
Shackle, G.L.S. (1974), Keynesian Kaleidics, Edinburgh: Edinburgh University Press.
Exchange Rates
The central feature of the Post Keynesian approach to exchange rates is the
belief that capital flows play an active, autonomous role in the economy.
This view is key in terms of both their explanation of currency price determination and their policy prescriptions; it is also what most distinguishes
Post Keynesian scholarship from orthodox. This entry begins with a review
of the latter.
Mainstream theories of exchange rates, though there are several, ultimately agree that the underlying forces driving foreign currency prices are
the fundamentals. Unfortunately, little eort is expended in explaining these
determinants (see Harvey 2001). Definitions range from simple lists of
potential candidates to circular references to them as those variables suggested by economic theory. In reviewing this literature it becomes apparent that the unifying theme among the seemingly disparate approaches is
that the fundamentals represent that set of variables guaranteeing the ecient operation of the foreign currency market. Orthodoxy is therefore
assuming the optimality of the outcomes created by real-life foreign
exchange operations; to them the task at hand is to identify these undiscovered variables.
A second feature of the mainstream approach is their acceptance of Says
Law and consequent relegation of monetary factors to irrelevance. No
orthodox exchange rate model treats portfolio capital flows as anything but
transitory factors, serving merely to finance trade flows. If short-term
investment moves a currency price, this is merely a reflection of the fundamental factors in the economies in question. Thus, though capital is undeniably the largest factor in the balance of payments today, it can be safely
ignored. It is no more responsible for determining prices than the mechanism which axes stickers to cans of soup at the grocery store (or, to oer
a more modern analogy, resets the registers interpretation of the universal
price code symbol).
As a result of these dispositions, orthodox exchange rate theory focuses
almost exclusively on trade flows as the determinant of currency prices.
Note that if currency prices are indeed driven by trade flows, trade imbalances must represent an excess demand for the money of the surplus nation.
A corollary to the orthodox approach is therefore that balanced trade can
be expected to prevail over the long run.
132
Exchange rates
Empirical tests of mainstream exchange rate models have fared very

poorly, even by the admission of supporters. Trade does not tend towards
balance, capital flows appear to have enormous influence on currency
markets, and exchange rates have been far too volatile for their only determinants to have been fundamental in origin. This has led to considerable
questioning within the orthodox school, the basic result of which has been
the conclusion that short-term currency movements may be irrational and
therefore non-economic; economists time, they conclude, is best spent
focusing on long-term movements. Note the striking similarity of this position to that of the classicals criticized by Keynes in The General Theory.
The Post Keynesian view is driven by their rejection of Says Law (due to
Keynesian uncertainty) and the consequent belief that finance plays a substantive role in determining output and employment. Within this context,
Post Keynesians argue that currency prices are a function of international
investors portfolio decisions. Moneys of nations whose assets are in greatest demand will tend to appreciate; those in least demand will depreciate.
While the value of international assets may be related to trade flows in some
indirect manner (which would tend to return us to the orthodox position),
it is primarily financial considerations that underlie agents decision
making. Greatest among these is the potential for capital gain from selling
an asset in the future. Expectations therefore play a central role in the
Post Keynesian explanation. However, unlike the rational expectations
employed by the orthodox school, which simply passively observe and
predict outcomes (the latter being generated by the fundamentals), those
posited in the Post Keynesian view create the objective variable. If market
participants expect (in aggregate) that European assets will become more
valuable, investors will enter the market to purchase those assets and thus
cause their (and the euros) appreciation. The trick to earning profits
becomes guessing the behaviour of the rest of the market.
How do agents form their expectations? That, of course, is the key question. It is also a very dicult one. In their search for a realistic formulation
of agents expectation-formation processes Post Keynesians have turned in
part to other disciplines. The work of the psychologists Daniel Kahneman
and Amos Tversky has provided an excellent starting-point. In their view,
people make decisions on the basis of simple heuristics, the most important
of which are availability, representativeness and anchoring. Availability is
used to estimate frequency or likelihood. In general, the more available
something is in memory (either through imagination or recalling past
instances), the more frequent or likely that event is deemed to be.
Representativeness is most useful when the decision maker is concerned
with the probability that object A belongs to class B (for example, the likelihood that event A is the result of process B, or that process B will create
Exchange rates 133

event A). The simple rule of thumb is, the more A resembles B, the more
likely it is that it belongs to class B. Anchoring occurs when the individual
must make an estimate. Psychologists have discovered that, regardless of
the procedure used to arrive at an initial estimate, people tend to anchor
any subsequent revisions to it.
Among the many implications of these rules of thumb are that agents
overemphasize the importance of events that are more recent or dramatic
(leading to overreaction in financial markets); they tend to place undue
confidence in making decisions based on scant information (which, combined with the Post Keynesian characterization of decision making in an
environment of uncertainty, tends to create volatility); and they will tend
to accept the current rate as the standard against which to consider future
movements (as in Keyness concept of convention). There is furthermore
a great deal of support from the psychological literature for the existence
of bandwagon eects in financial markets. This is key because it helps to
explain what appear to be self-sustaining trends and the existence and
popularity of technical analysis (which is generally based on the premise
that emerging trends will continue). Orthodoxy, in relying on the fundamentals approach, has discounted the importance of both bandwagons
and trading rules. However, empirical analysis has found them to be
important in explaining short-term currency movements (the time
horizon that orthodoxy has all but abandoned). The multi-disciplinary
approach of Post Keynesian economics also provides theoretical underpinnings for other observed phenomena, such as cash-in eects (the tendency of agents to sell appreciating assets) and currency-desk mandated
trading limits for dealers (as a defence against risk-taking behaviour in the
face of losses).
In summary, the Post Keynesian view of exchange rates is based on the
empirical fact that currency prices are driven by short-term capital flows.
Those flows are in turn a function of agents expectations, which are best
modelled using a combination of standard Post Keynesian tools (Keyness
concepts of convention, uncertainty and so on) and some borrowed from
psychology. The picture that emerges is not one of a market characterized
by stability, eciency and optimality (all benevolently guided by the invisible hand of the fundamentals), but of an institution where agents imperfectly considered actions create currency prices. Those actions may be
marked by stability for long periods of time (though with whipsaw patterns
created by the interactions of the cash-in and bandwagon eects) as agents
rely on convention to anchor to stable levels or rates of change; but because
they are subject to availability and bandwagon eects they are apt to rapid
revision in the face of salient events (even when those events may seem
inconsequential to cooler heads).
134
Exchange rates
Currency markets aect world welfare in a number of ways. First, international commerce (especially trade) is discouraged by the volatility of
exchange rates. Although one could correctly argue that this instability has
hardly brought import and export activity to a halt, it has at the very least
caused a shift of world resources away from productive activities and into
pecuniary ones. Ironically, the change in focus entailed by that shift has no
doubt exacerbated the problem by increasing the size and importance of
international capital.
Second, if one of the goals of an international monetary system is a tendency to balanced trade, then our current arrangements do not and cannot
deliver. Simply put, one price cannot except by coincidence clear two
markets. Were there no capital flows in the world then (assuming minimal
government intervention) trade imbalances would represent market disequilibria. Current account deficits and surpluses would soon be eliminated.
Alternatively, were capital flows very small as compared to trade flows, then
agents in the former would have a vested interest in tracking the latter; they
would operate to anticipate current account activities and to oer liquidity
on a timely basis. However, given that the overwhelming majority of international commercial activity takes the form of portfolio investment, there is no
reason to believe that events in the worlds current account are of more than
minor interest to the agents acting in the capital account. Even when a nation
is heavily indebted and one might expect a massive depreciation/devaluation
to take place to correct an accompanying trade deficit, investors have rightly
come to anticipate that the more normal course of action will be emergency
loans and grants that allow a continuation of the status quo. There is no automatic stabilizer operating in the post-Bretton Woods world.
As a consequence of their understanding of the foreign exchange
market, the most common recommendation of Post Keynesian economists
with respect to the organization of the international economy is that we
control the flow of capital. This notion is hardly new. Keynes argued at the
Bretton Woods conference that nations must reserve the right to control all
capital movements. His recommendation was not entirely ignored, but the
eorts made were half-hearted and, in the end, easily circumvented by
investors. Of course, addressing this shortcoming has now been complicated by the laissez-faire attitude that prevails among policy makers and
academics.
But reforming the international monetary system will require more than
simply slowing the rate of international portfolio investment flows. In recognition of this fact, Paul Davidson has recommended a comprehensive
plan that addresses both currency and broader macro issues. At its core is
a recognition that Says Law does not operate in the real world and that we
must therefore undertake policy to generate full employment. The system
Expectations
135
would feature a fixed exchange rate (as a means of making international

transactions prices more predictable) and provisions to control capital
flows (so that those fixed rates could be defended). Furthermore, it would
place the burden of resolving trade imbalances on the surplus country and
not the deficit country. This is logical, according to Davidson, because in a
less-than-full-employment world where Says Law does not hold, surplus
nations are antisocial drains on world employment. This proposed system
will thereby avoid the inherent deflationary biases of those based on orthodox understandings of the currency market.
J T. H
See also:
Bretton Woods; Economic Policy; Expectations; Globalization; International Economics;
Says Law; Tobin Tax.
Bibliography
Davidson, Paul (1998), Volatile financial markets and the speculator, Economic Issues, 3 (2),
118.
Harvey, John T. (1993), Daily exchange rate variance, Journal of Post Keynesian Economics,
15 (4), 51540.
Harvey, John T. (1999), Exchange rates: volatility and misalignment in the post-Bretton
Woods era, in Johan Deprez and John T. Harvey (eds), Foundations of International
Economics: Post Keynesian Perspectives, London: Routledge, pp. 200211.
Harvey, John T. (2001), Exchange rate theory and The fundamentals, Journal of Post
Schulmeister, Stephan (1987), An essay on exchange rate dynamics, Research Unit Labour
Market and Employment Discussion Paper 878, Berlin: Wissenschaftzentrum Berlin fr
Sozialforschung.
Expectations
One of the intrinsic properties of our world is that the future is uncertain.
This uncertainty aects the way decisions are made because economic
actors (entrepreneurs, bankers, employees, government) have to define their
priorities by relying on their expectations about the future so as to anticipate it (that is, to act in advance of what they think will happen). In The
General Theory, Keynes explains how and why expectations influence the
current and future states of the economic system and how expectations are
formed. He was followed by many dierent authors, including Joan V.
Robinson, Nicholas Kaldor, Roy F. Harrod, George L.S. Shackle, Evsey D.
Domar, Hyman P. Minsky, Paul Davidson and Jan A. Kregel, who have
clarified and developed Keyness analysis.
The first problem when dealing with expectations is to look at how and
136
Expectations
why they matter. Because we live in a monetary production economy, the

most influential expectations are those of entrepreneurs and of the banking
and financial communities. Keynes distinguishes between short-term
expectations, long-term expectations and confidence in these expectations.
The first consist of expectations of costs and sale proceeds induced by a
certain level of production, for a given level of capital equipment (new
capital assets are not available for production yet). The aim of this calculation is to find progressively the point of eective demand, that is to say, the
point at which production is expected to give the highest reasonable profits
(the highest profits possible without jeopardizing the long-term viability of
the firm). This point represents the current equilibrium position (by which
Post Keynesians mean state of rest and not market clearing) induced by a
given state of long-term expectation. The short term refers here to the
shortest interval after which the firm is free to reverse its decision as to how
much employment to oer (Keynes 1936, p. 47). These short-term expectations are gradually revised in the light of current economic results.
However, to simplify the analysis, Keynes considers that the theory of
eective demand is substantially the same if we assume that short-term
expectations are always fulfilled (Keynes 1937b, p. 181). Thus, one can consider that entrepreneurs do not, as a rule, make wildly wrong forecasts of
the equilibrium position (p. 182) so that if we suppose a state of expectation to continue for a sucient length of time for the eect on employment
to have worked itself out . . . completely . . . the steady level of employment
thus attained may be called the long-period employment corresponding to
that state of expectation (Keynes 1936, p. 48). Then, with this simplification, the point of eective demand is the actual level of production.
What really matters for the economic system are the past and current
states of long-term expectations (Kregel 1976). These concern the future
net cash flows provided by old and new capital assets. The past states of
long-term expectation are reflected in the current amount and composition
of the capital equipment. In the simple Keynesian economic model, the
current state of long-term expectation (named E in a draft of The General
Theory) determines the expected level of expenditures for investment (I)
and consumption (C). Indeed, E is a key variable for the marginal propensity to consume (and so C), but also, and mainly, for the marginal eciency
of capital and the interest rate (and so I). Therefore, the aggregate demand
curve (D) and aggregate supply curve (Z) are drawn for a given E.
Minsky (1975) elegantly showed how the state of long-term expectation
aects the level of investment (that is, the production of new capital assets)
and, via the multiplier, the level of employment. Thus, for two dierent
states of long-term expectation with short-term expectations realized, see
Figures 9 and 10.
Expectations
Pk , PI
n
Pk
PI
n
IF
Figure 9
I1
I2
Determination of the level of investment
Z, D
Z'
D'
De(I2)
D
De(I1)
N(I1)
Figure 10
N(I2)
Determination of the level of employment
N*
137
138
Expectations
Pk is the demand price of capital assets. It is the marketable price of existing capital assets. This price depends on the entrepreneurs (or borrowers)
risk, which is the entrepreneurs sentiment about the viability of his project
(or, at the macroeconomic level, entrepreneurs opinion about the current
and future states of the economy). Thus, Pk is determined by the expectations entrepreneurs make about the future net cash flows (n) generated by
a certain level of investment. Pl is the supply price of capital assets; the
price to pay to get new capital assets. This price is fixed by the producers of
capital assets out of a mark-up over costs. When investment is not completely self-financed (I IF), the supply price depends on the lenders risk,
which is the lenders opinion about the creditworthiness of a borrower.
Then PI is influenced by the expectations of lending institutions concerning the future net cash flows provided by capital assets. Indeed, these expectations greatly influence the cost of external financing (interest rates; prices
of stocks).
As long as economic conditions are favourable (Pk Pl), the equalization
of the two prices determines I. The more optimistic both entrepreneurs and
lending institutions are, the higher the level of investment is, leading to a
higher level of eective demand (and so of production). However, it is sucient for the state of long-term expectation of one community (entrepreneurs, financial analysts or banks) to be degraded to decrease the level of
investment. This depressive eect may grow because entrepreneurs and
lending institutions are influenced by each other while determining their
long-term expectations. Thus, more pessimistic (optimistic) expectations
from one community can lead to more pessimistic (optimistic) in the other
one: Pk and Pl are related (Keynes 1936, p. 145).
Actually the reality is worse than that because the two prices do not
depend only on the lending institutions expectations (for Pl) and entrepreneurs expectations (for Pk). Indeed, the expectations of speculators are
also very important. Kaldor (1939) shows in great detail how speculation
may generate economic instability. Speculators make portfolio arbitrages
to earn short-term capital gains, so they are not interested in future rents
provided by a capital asset during its entire life. For speculators, buying, for
example, financial assets is a game that consists in trying to anticipate what
the main opinion of the financial community will be in the short term. This
is the famous beauty contest situation described by Keynes in chapter 12 of
The General Theory. Each judge of the contest is asked not to try to find
the most beautiful woman (that is, the most economically viable capital
assets), but the woman that other judges consider the most beautiful (that
is, assets that others will buy in the near future). The problem is that this
kind of behaviour has a tendency to become generalized in financial
markets; it is possible to make quick capital gains and, by not adopting this
Expectations
139
behaviour, one may make capital losses. Thus, speculators can lead both
lending institutions and entrepreneurs to adopt a speculative behaviour
(forecasting the psychology of the market) instead of an enterprise
behaviour (that is, act in expectation of future income streams provided by
the productive use of capital assets). Therefore, the state of long-term
expectation, which is already very fragile because it depends on forecasts of
economic variables in the long run about which we know very little (wages,
interest rates, tastes of consumers, degree of competition and other variables), is still more precarious because it depends on fads and fashions in
the financial markets. This, however, does not mean that the state of longterm expectation is completely unstable. It means only that this state can
change very abruptly for insignificant or purely cyclical reasons. These
changes are largely independent of the realization or not of short-term
expectations, because it is of the nature of long-term expectations that they
cannot be checked at short intervals in the light of realized results (Keynes
1936, p. 51).
However, for Keynes and Post Keynesians the real problem is elsewhere
(Kregel 1976). What really matters for employment is neither the fulfilment
or not of short-term expectations, nor the instability of long-term expectations. It is, instead, the level of the latter that is important: high unemployment (N* Ni) results from an insucient level of eective demand (De).
Stated alternatively, because economic actors are too pessimistic, their
spending is too low to implement a level of production that is consistent
with full employment. This shows that the conditions of equilibrium are
endogenously determined by economic actors actions, which are based on
their expectations. There is no predefined state that the economy can reach;
this state is created and modified by economic actors anticipations (Kregel
1986). At the macroeconomic level, economic actors are responsible for
their own economic situation. Then, pessimism may be so high that economic actors cannot solve their problems by themselves, or, economic
actors can be too optimistic and indulge in dangerous behaviours leading
to doubtful investments and financial weakness. Indeed, during a boom
the popular estimation of the magnitude of both . . . borrowers risk and
lenders risk, is apt to become unusually and imprudently low (Keynes
1936, p. 145). The government should thus do everything to limit and to
regulate the economic impact of pessimism and optimism. Indeed, this
economic actor has the financial and technical ability to anticipate the
macroeconomic consequences of the current anticipations of private
agents. Entrepreneurs are only concerned with expectations of profits of
their own activities; and it is not their job to take into account the macroeconomic consequences of their own decisions.
The preceding has shown how expectations influence the current state of
140
Expectations
the economy (determination of the level of production and of the current

level of capital equipment) and why they are important (insucient eective demand or unsustainable economic patterns). It is now necessary to
look at how the state of expectation is formed to better understand how it
evolves. Here the confidence with which we make this forecast (Keynes
1936, p. 148) becomes important.
Once again it is necessary to make a distinction between short-term
expectations and long-term expectations. Concerning the former, Keynes
agrees that they can quite easily be formed by using probability calculus. To
find the eective scale of production, an entrepreneur will make several
hypothetical expectations held with varying degrees of probability and definiteness (p. 24 n. 3). The constant overlapping between short-term expectations and current results (p. 50) makes this probabilistic calculation quite
easy because it is based on a routine process.
If probabilities can be used more or less easily for short-term expectations, long-term expectations cannot be based mainly on this kind of
method of decision. The mathematical expectation is, at best, an element
among others in the process of decision. Indeed, to apply probability calculus with confidence and to base his or her actions only on this method of
decision, an economic actor has to assume that the economic system is
ergodic. This means that the properties of the system in which decisions are
made are not modified by these decisions (Davidson 1991). However,
investment spending is a crucial decision (Shackle 1955) because it leads to
irreversible qualitative changes in the economic system. Thus, future possibilities and properties of the economic system evolve as new technologies
and methods of production are introduced. There is another important
reason why long-term expectations cannot rely essentially on probabilities:
these expectations depend on factors that we know little about, so that the
state of confidence plays a dominant role in the formation and change of
long-term expectations. This implies that, contrary to probability calculus,
probability and confidence do not necessarily move in the same way and
have to be clearly separated. It is not because more information is available
that confidence is increased; events, such as crises, can radically diminish
the confidence with which views of the world are held (Minsky 1975, p. 65).
Thus, the state of confidence, as they term it, is a matter to which practical men always pay the closest and most anxious attention (Keynes 1936,
p. 148). To evaluate this state of confidence, economic actors refer to the
prevailing convention concerning the present and future states of the
economy. This means that, to take crucial decisions, they rely heavily on
the past and current economic situations, and that they judge the current
opinion of the majority as the best (Keynes 1937a). However, entrepreneurs can also go against the convention in place and let their instinct dom-
Expectations
141
inate their decisions. Long-term expectations then depend on dierent elements that are related but have little to do with probability (Dequesh,
1999): animal spirits, creativity and uncertainty perception are three of
them. It follows from this that conventions are fragile and subject to sudden
changes. However, they are usually stable enough for entrepreneurs whose
optimism and animal spirits push them to invest. Moreover, this uncertainty about the future is less an obstacle than a stimulus to investment.
Uncertainty leaves the system open to the imagination of entrepreneurs
regarding profit opportunities. If investment depended on nothing but a
mathematical expectation, enterprise [would] fade and die (Keynes 1936,
p. 162).
T
See also:
Non-ergodicity; Treatise on Probability; Uncertainty.
References
Davidson, Paul (1991), Is probability theory relevant for uncertainty?, Journal of Economic
Perspectives, 5 (1), 12943.
Dequesh, David (1999), Expectations and confidence under uncertainty, Journal of Post
Kaldor, Nicholas (1939), Speculation and economic activity, Review of Economic Studies, 7
(1), 127. Reprinted and revised in N. Kaldor, Essays on Economic Stability and Growth,
London: Duckworth, 1960, pp. 1758.
Keynes, John M. (1936 [1973]), The General Theory of Employment, Interest and Money.
Reprinted in The Collected Writings of John Maynard Keynes, Vol. 7, London: Macmillan
for the Royal Economic Society.
Keynes, John M. (1937a) [1973]), The general theory of employment, Quarterly Journal of
Economics, 51 (2), 20923. Reprinted in The Collected Writings of John Maynard Keynes,
Vol. 14, London: Macmillan for the Royal Economic Society, pp. 10923.
Keynes, John M. (1937b [1973]), Ex post and ex ante, 1937 lecture notes. Reprinted in The
Collected Writings of John Maynard Keynes, Vol. 14, London: Macmillan for the Royal
Economic Society, pp. 17983.
Kregel, Jan A. (1976), Economic methodology in the face of uncertainty: the modeling
methods of Keynes and the Post-Keynesians, Economic Journal, 85 (342), 20925.
Kregel, Jan A. (1986), Conceptions of equilibrium: the logic of choice and the logic of production, in I. Kirzner (ed.), Subjectivism, Intelligibility, and Economic Understanding, New
York: New York University Press, pp. 15770. Reprinted in P. Boettke and D. Prychitko
(eds.), Market Process Theories, Volume 2: Heterodox Approaches, Cheltenham, UK and
Northampton, MA, USA, Edward Elgar, 1998, pp. 89102.
Minsky, Hyman P. (1975), John Maynard Keynes, London: Macmillan.
Shackle, George L.S. (1955), Uncertainty in Economics, Cambridge: Cambridge University
Press.
Finance Motive
The finance motive for holding money was introduced by J.M. Keynes
(1937a, 1937b, 1939) one year after the publication of the General Theory
(1936) and has been since then the object of a long and lively debate.
However, the concept and the term have remained strictly related to Keyness
writings and never became a part of the common economic language.
In the General Theory, Keynes had concentrated his attention on money
as a store of wealth, neglecting the analysis of money as an intermediary
of exchanges. The reason for so doing was in part a purely analytical one.
Only if considered as a stock does money become an observable and measurable variable; on the other hand, in any single instant of time, the whole
of the money stock must be present in the liquid holdings of some agent or
other. A second reason, possibly the main one for concentrating on money
as a stock, was that Keynes considered the demand for liquid stocks to be
at the origin of prolonged failures of aggregate demand, the main phenomenon he wanted to explain.
Severe criticisms of the Keynesian treatment of money in the General
Theory were put forward by as authoritative an economist as D.H.
Robertson, who accused Keynes of being so taken up with the fact that
people sometimes acquire money in order to hold it, that he had apparently
all but entirely forgotten the more familiar fact that they often acquire it in
order to use it (1940, p. 12). Robertsons remarks convinced Keynes that his
model had to allow not only for money lying idle in somebodys holdings
but also for money moving from the holdings of one agent to the holdings
of another one, thus allowing the exchange of goods and services to take
place.
In three articles published between 1937 and 1939, Keynes answered his
critics and completed his model by a simple construction containing a
description of the whole process of money creation. Any agent wanting
finance for his business will ask for credit, usually (but not necessarily) from
a bank. If liquidity is supplied by a bank, the money stock is increased. If
it is supplied by some other agent in possession of idle money and being
willing to lend it, the money stock does not change and the velocity of circulation is increased. As soon as the bank (or some other agent) has granted
him the required credit, the agent is in possession of a liquid sum. It is clear
that the agent is now holding money not for the sake of keeping it idle but
in order to spend it. In fact he will only hold his money balance for the short
time covering, in Keyness own words, the interval between planning and
142
Finance motive 143

execution [of expenditure] (1937b, p. 663 [1973, p. 216]). In this case,
according to Keynes, the motive behind the demand for money is not a
transaction, or a precautionary, or a speculative motive, but a fourth
motive, named by Keynes the finance motive for holding money. As soon as
it is spent, money initially held as finance enters the money holdings of
some other agent (wage-earners, suppliers of intermediate goods) and
becomes, as the case may be, one of the more familiar transactions, precautionary, or speculative balances.
By his analytical construction, Keynes was trying to reconcile the
demand for money as an intermediary of exchange, historically the first
and more intuitive concept of money, and the demand for money as a
demand for a store of wealth, the basic definition around which he had
built the analysis of the General Theory.
Simple as the Keynesian construction may be, it has given rise to a
number of misunderstandings. Most interpreters of Keynes believe that
finance is only required when firms are facing an increase in output and that
a stationary level of output is somehow self-financed (Chick 1983,
pp. 198200). In their view, the finance motive for holding money is strictly
connected to the transactions motive, the only dierence between the two
being that while the first one defines the demand for liquid balances as a
function of current income, the second one defines an increase in the
demand for money as a function of an increase in the expected level of
income.
Keynes himself may have induced a similar interpretation by his insisting on the fact that the use of finance is what makes possible an increase in
production (1937a, p. 247 [1973, p. 209]; 1937b, p. 668 [1973, p. 222]).
However this is clearly wrong. What is true is that, in a stationary economy,
a constant level of output may require a constant volume of finance (if the
velocity of circulation is constant). In that case, finance, as Keynes himself
said, may become a constant revolving fund, used again and again. However, in a monetary economy, where money is the only means of payment
and the banks are the only producers of money, no level of output can be
obtained if it is not duly financed by the banks, and the revolving fund itself
has to be supplied by the banks.
The same idea that the finance motive should be assimilated to the transactions motive for holding money has given rise to the suggestion that
finance, being an advance provision of cash for investment, might have
nothing to do with speculation. If this were true, introducing the finance
motive might mean neglecting the conflict between finance and industry, a
fundamental feature of the General Theory. However, this does not seem to
be strictly correct. Keynes himself, when describing the demand for finance,
explicitly mentions the fact that finance covers equally the use of the
144
Finance motive
revolving pool of funds to finance . . . (e.g.) an increased turnover on the

stock exchange (1939, p. 573 [1973, p. 283]). The presence of finance in the
model in no way rules out conflict between finance and industry. It rather
introduces into the model a new, and possibly more interesting, kind of
conflict, namely the one between bankers and entrepreneurs. In the General
Theory Keynes had given space to the conflict between entrepreneurs and
rentiers. In his later articles his purpose is to introduce what he names the
power of the banks (1937a, p. 248 [1973, p. 211).
A second misinterpretation, common to most authors dealing with
finance, is to consider finance as only needed when investment expenditure
is involved (Asimakopulos 1983). In fact, finance is required for any kind
of output, and no distinction can be made between production of consumer goods and production of capital goods. Keyness statements in this
direction are as repeated as they are clear: The production of consumption
goods requires the prior provision of funds just as much as does the production of capital goods (Keynes 1939, p. 572 [1973, p. 282]; see also
Keynes 1937a, p. 247 [1973, p. 208]; 1937b, p. 667 [1973, p. 221]; 1939, p.
573 [1973, p. 283]).
The prevailing confusion between finance and investment is responsible
for the peculiar statement that, whenever the level of investment is increased,
firms can repay their bank debt only after the multiplier process has fully
worked itself out (Cesaroni 2001). It is of course quite true that it is only
when the multiplier process has come to an end that ex ante (or voluntary)
savings are again equal to investment. But this has nothing to do with
finance. Once investment is executed, an increase in income equal to it is
created. If liquidity preference is stable and government securities are absent,
the new income will be entirely spent either on the commodity market or on
securities issued by private firms. Therefore, even after the very first round of
expenditure, the firms will be fully able to repay their bank debt.
The debate on Keyness finance motive shows that 65 years after the publication of the General Theory, confusion still persists between prior
finance, needed for any kind of production (the problem Keynes was trying
to analyse), and the totally dierent problem of how an adequate supply of
saving is generated in order to bring saving and investment to equality. The
first one is a problem concerning the credit market; the second one is the
problem of ensuring equilibrium in the commodity market. The fundamental distinction between finance and saving that Keynes was trying to make
does not seem to have been absorbed, and his conclusion that the investment market can become congested through a shortage of cash. It can
never become congested through a shortage of saving (1937b, p. 669 [1973,
p. 222]) seems to be still ignored.
A G
Financial instability hypothesis
145
See also:
Banking; Circuit Theory; Investment; Liquidity Preference; Money; Multiplier.
Bibliography
Asimakopulos, A. (1983), Kalecki and Keynes on finance, investment, and saving,
Cambridge Journal of Economics, 7 (34), 22133.
Cesaroni, G. (2001), The finance motive, the Keynesian theory of the rate of interest and the
investment multiplier, European Journal of the History of Economic Thought, 8 (1), 5874.
Chick, V. (1983), Macroeconomics after Keynes, Oxford: Philip Allan.
Macmillan.
Keynes, J.M. (1937a), Alternative theories of the rate of interest, Economic Journal, 47 (186),
24152 (reprinted in Keynes 1973, pp. 20115).
Keynes, J.M. (1937b), The ex-ante theory of the rate of interest, Economic Journal, 47 (188),
6639 (reprinted in Keynes 1973, pp. 21523).
Keynes, J.M. (1939), The process of capital formation, Economic Journal, 49 (195), 56974
(reprinted in Keynes 1973, pp. 27885).
Keynes, J.M. (1973), The Collected Writings of John Maynard Keynes. Volume XIV: The
General Theory and After. Part II: Defence and Development, London: Macmillan for the
Kregel, J. (1986), Keynes and finance. From the Treatise to the General Theory, (mimeographed) (printed in Italian: Il finanziamento in Keynes. Dal Trattato alla Teoria Generale,
in M. Messori (ed.), Moneta e produzione, Turin: Einaudi, 1988, pp. 5971).
Robertson, D.H. (1940), Mr Keynes and the rate of interest, in D.H. Robertson, Essays in
Monetary Theory, London: P.S. King & Son, pp. 138.
Financial Instability Hypothesis

Hyman Minsky had a long and distinguished career that spanned almost
four decades, during which he developed a number of key insights into the
workings of modern financial economies. Minskys work is complex and
rich, and attempts to model the real world in which financial institutions
play a key role. Minskys work has had a definite influence, not only on Post
Keynesians, but also on institutionalists and Marxists. Recently, however,
New Keynesians have also showed a keen interest in his work.
For Minsky, orthodox theory is best described as a village market where
bartering one good for another is the principal economic activity. A capitalist economy, in contradistinction, is much closer to a Wall Street system
where agents, businesspeople and bankers deal with investment financing
and capital assets. It is within this setting that Minskys financial instability
hypothesis is developed. It deals with a capitalist economy of production,
in Keyness sense, where finance and financial institutions play a key and
decisive role.
Among Minskys many important contributions, the financial instability
hypothesis remains his most important. It explains the inherent cyclical
nature of modern financial economies, and how economic booms can sow
146
the seeds for an eventual downturn, that is, how stability breeds instability.
Capitalist economies cannot be studied without referring to their monetary
and financial nature.
Minskys work on financial instability incorporates aspects of Keynes,
Michal- Kalecki and Irving Fisher. First, Minskys work is set within an
environment of Keynesian uncertainty. Businesses and banks operate in an
uncertain environment in Keyness sense, and hence their expectations and
decisions are made in a world devoid of reliable knowledge. Second,
Kaleckis principle of increasing risk is another central component of
Minskys work, according to which firms and the macro economy become
more fragile as their level of debt increases. As it expands, the economy
becomes increasingly fragile. In this sense, financial cycles are endogenous.
Finally, as the economy collapses, deflation may ensue, implying that debt
incurred during the expansionary phase of the cycle may not be reimbursed. This may then lead to debt-induced bankruptcies and a deepening
recession. However, expansionary fiscal policy and a central bank acting as
a lender of last resort may help in limiting the scope of the recession.
For Minsky, the early stages of an economic cycle are best described as
periods of caution, as agents remember the last phases of the previous
cycle. Coming out of a recession, firms tend to undertake safe investment
projects where the expected revenues exceed the necessary debt repayments.
Agents liability structures are very liquid and the debt/equity ratio of firms
is relatively low, or at least within respectable or acceptable levels. Firms
expect good returns and expectations are generally fulfilled. In this stage,
the economy is in a tranquil phase of hedge finance. Firms tend to finance
their investment initially through retained earnings, or at least internal
financing is much greater than external financing.
As the boom continues, however, firms decide to undertake additional
investment. Minsky (1982, pp. 12024) refers to this phase as economic
euphoria. The optimism is fuelled by growth, and is shared by banks
(p. 121). This is a key element. As firms invest more than their retained
earnings allow, they will seek access to bank credit. Provided banks are as
optimistic as firms, they will finance new investment. Simultaneously, asset
prices start rising as speculators enter the market. These Ponzi financiers
tend to borrow heavily to purchase assets in the hope of selling them at
higher prices.
As both firms and speculators become more indebted and less liquid, interest rates start to rise, as rates are positively correlated with debt/equity ratios.
It is the illiquidity of both banks and firms that fuels the rise in interest rates.
This rise of interest rates places the economy at risk, as firms may not be able
to meet their debt commitments. Refinancing existing debt is made at a
higher rate of interest, implying that cash outflows are greater than cash
147
inflows. Debt burdens are increasing. Higher interest rates and less-liquid
balance sheets also imply growing fragility of the banking system. This is
when the economy moves into a situation of Ponzi finance. Financial
euphoria slowly leads to financial panic, and a crisis may be at hand. At this
point, asset prices and gross profits collapse; investment falls or even stops.
The economic boom is now replaced by an economic downturn.
The degree to which the economy spirals downward will depend largely
on the role of prices, but also on fiscal and monetary policy. If price inflation is high, firms revenues may be sucient to permit them to honour part
of their debt commitments. If price inflation is low, however, accumulated
debt will be too much of a burden, and the economy will continue to spiral
downward.
Minskys work on financial instability carries important policy implications in the Keynesian tradition. Since it discusses the inherent tendency for
economies to go from booms to busts, it addresses the specific roles of fiscal
and monetary policies in constraining the dynamic nature of capitalist
economies. In fact, since Minskys work relies on developed modern institutions, it can be used to explain why large-scale depressions have not
occurred since the 1930s. Since the public sector was small, fiscal policy
could not have prevented the Great Depression. It is in this sense that
Minskys work is institutionally sensitive.
Today, however, the story is much dierent and governments are active
and important players in the real world. Fiscal policy can have an important role in preventing further economic malaise, as fiscal deficits can
translate into larger gross profits, enabling firms to honour their cash commitments on outstanding debt. Moreover, fiscal deficits may also limit the
extent to which debt deflation occurs. According to Minsky (1982, p. xx):
A cumulative debt deflation process that depends on a fall of profits for its
realization is quickly halted when government is so big that the deficit
explodes when income falls. Furthermore, the central bank can have an
important role in preventing runs on banks. It does so by expanding the
monetary base to allow sucient liquidity. It can also relax certain regulatory rules, like reserve requirements. By doing so, a liquidity crisis can be
avoided. Both policies can help in preventing continued deterioration of
money profits, which are important for debt validation and asset prices.
Minskys reliance on modern institutions explains in fact why It has not
happened again.
In Minskys world, capitalism is not a system that tends naturally to
stability. It is fraught with chaotic episodes and tendencies to periodic
booms and slumps. This is characteristic of financial and monetary economies in an uncertain world. This does not mean that a depression cannot
happen again. Governments and central banks may choose not to act.
148
Moreover, sound fiscal and monetary policies do not eliminate the financial phases of economic cycles, as these are endogenous to the cycle. Policy
cannot avert the existence of Ponzi speculators.
Minskys financial instability hypothesis has raised some concern among
Post Keynesians. Key to this criticism is the fact that Minskys analysis of
financial fragility is essentially based on the microeconomic behaviour of
the bank and the firm and is devoid of macroeconomic significance. This
has led Lavoie and Seccareccia (2001) to question the missing macroeconomic link.
Minskys analysis of the notion that economic expansion leads to higher
debt/equity ratios that translate automatically into higher interest rates
may be applicable to the individual firm or bank. As their debt/equity ratios
increase, banks may perceive them as riskier and may charge a higher rate
of interest to cover the higher risk. His analysis, however, may not necessarily hold for the macro economy. In other words, as the economy
expands, it may not necessarily become more fragile or riskier, and there is
no reason why rates of interest need to increase, especially in an environment of endogenous money with exogenous rates of interest. These are set
by the central bank.
Minsky, in fact, provides only one example of his financial fragility
hypothesis in a macroeconomic setting. It can be found in an early article
in the American Economic Review (Minsky 1957). The only problem is that
the argument is set within the loanable funds approach (Lavoie 1996;
Lavoie and Seccareccia 2001; Rochon 1999), which would explain why
interest rates automatically increase during expansions.
Furthermore, Minskys early analysis is silent on Kaleckis profit equations (although they figure in his later writings: see Minsky 1977). Had
Minsky taken note of these equations, he would perhaps have realized that
debt/equity ratios might not rise during expansions, which would then
imply that the economy does not necessarily become increasingly fragile.
As Lavoie and Seccareccia (2001, p. 84) argue, There is a missing link.
Minsky does not provide any rationale to justify his rising leverage ratio
thesis at the macroeconomic level.
L-P R
See also:
Banking; Business Cycles; Central Banks; Fiscal Policy; Kaleckian Economics; Liquidity
Preference; Monetary Policy; Rate of Interest; Uncertainty.
Bibliography
Bellofiore, R. and P. Ferri (eds) (2001), Financial Keynesianism and Market Instability: The
Economic Legacy of Hyman Minsky, two vols, Cheltenham, UK and Northampton, MA,
USA: Edward Elgar.
Fiscal policy
149
Lavoie, M. (1996), Horizontalism, structuralism, liquidity preference and the principle of

increasing risk, Scottish Journal of Political Economy, 43 (3), 275300.
Lavoie, M. and M. Seccareccia (2001), Minskys financial fragility hypothesis: a missing
macroeconomic link?, in R. Bellofiore and P. Ferri (eds), Financial Fragility and Investment
in the Capitalist Economy: The Economic Legacy of Hyman Minsky, Volume II,
Cheltenham,UK and Northampton, MA, USA: Edward Elgar, pp. 7696.
Minsky, H. (1957), Monetary systems and accelerator models, American Economic Review,
47 (6), 85983.
Minsky, H. (1977), The financial instability hypothesis: an interpretation of Keynes and an
alternative to standard theory, Nebraska Journal of Economics and Business, 16 (1), 516.
Minsky, H. (1982), Can It Happen Again? Essays on Instability and Finance, Armonk, NY:
M.E. Sharpe.
Minsky, H. (1986), Stabilizing an Unstable Economy, New Haven: Yale University Press.
Nasika, E. (2000), Finance, Investment and Economic Fluctuation: An Analysis in the Tradition
of Hyman P. Minsky, Cheltenham,UK and Northampton, MA, USA: Edward Elgar.
Rochon, L.-P. (1999), Credit, Money and Production: An Alternative Post-Keynesian Approach,
Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Fiscal Policy
Fiscal policy is concerned with the economy-wide eects of government
expenditure and revenue raising. Post Keynesians believe that there is no
endogenous mechanism in a capitalist economy which will ensure that economic activity tends to full employment, even in the long run. Hence, fiscal
policy is important as a major way in which the government can raise aggregate demand to the full-employment level. In a justly celebrated article,
Lerner (1943) argued that fiscal policy should not be based on sound
finance in which expenditure was balanced by revenue over a year or some
other arbitrary period. Instead fiscal policy should be based on functional
finance. Government expenditure and revenue should be determined so
that total expenditure in an economy is at the rate which will produce full
employment without inflation. This is to be done without any concern
about whether the resulting budget is in surplus or deficit.
With the neoclassical resurgence in the 1970s and succeeding decades,
sound finance was again adopted by many economists and policy makers
as the overriding guide to good fiscal policy. While Post Keynesian economists generally have a more complex theory of inflation than that implied
by Lerner in 1943, they share his rejection of sound finance. Much of the
Post Keynesian writing on fiscal policy is designed to counter neoclassical
arguments against budget deficits. Nevile (2000) contains a survey of these
neoclassical arguments and the counters to them.
Among the neoclassical arguments, crowding-out theory and the twin
deficits hypothesis are of particular importance, as each has had a substantial impact on actual policy making. Crowding-out theory maintains that
an increase in the deficit will cause a fall in private investment expenditure
150
Fiscal policy
of (almost) the same size as the rise in the deficit. If the government
borrows to finance the deficit this, it is argued, will force up interest rates,
reducing private investment. Moreover, even if the various multiplier eects
are such that economic activity increases, more money will be demanded
by the public to carry out this increased economic activity. They will try to
borrow this extra money, forcing up interest rates further until the increase
in gross domestic product is reversed.
An assumption underlying this crowding-out thesis is that the monetary
authorities are successful in maintaining a constant stock of money. Even
if the monetary authorities were successful in doing this, the analysis that
shows increased government expenditure leading to higher interest rates
also shows that any increase in private expenditure will lead to a rise in
interest rates. In this respect, expansionary fiscal policy is no dierent from
any other stimulus that might lift the economy out of recession. However,
even before financial deregulation the monetary authorities in developed
economies did not maintain a constant volume of money. Since financial
deregulation, the volume of money is endogenous. In eect those supporting crowding out in todays world of deregulated financial markets are
arguing that, whenever government expenditure increases, the central bank
actively tightens monetary policy to the extent necessary to reduce private
investment by an amount equal to all, or most of, the increase in public
expenditure.
There is one qualification that should be made to this conclusion: shortterm interest rates are the monetary policy instrument, but long-term interest rates may be more relevant to investment decisions in the private sector.
It is possible that large budget deficits might increase the spread between
short- and long-term interest rates, for example because they increase
expectations of inflation, so even if short-term interest rates were held constant long-term rates could rise, crowding out private investment. However,
there is no evidence that this happens, and more generally empirical studies
have found little evidence of a relationship between budget deficits and
interest rates. (See, for example, Nevile 2000, pp. 16061 and endnote 11.)
In addition, many Post Keynesians would deny that the link between interest rates and private sector investment is strong.
The second influential argument, the twin deficits hypothesis, maintains
that if a budget deficit is created or increased, the balance of payments
current account deficit will increase by a very similar amount so that all the
expansionary impact will go overseas through increased imports. The social
accounting identities ensure that this will happen if other things do not
change, but this proves nothing unless one has a theory to support the
implied ceteris paribus assumption. Supporters of the twin deficits hypothesis usually have no theoretical foundations for their arguments and those
Fiscal policy
151
that have been put forward hold only in very long-run equilibrium situations, making them largely irrelevant to anti-cyclical policy making.
Moreover, empirical evidence does not support the twin deficits hypothesis. For example, from 1990 to 1993 in G7 countries on average budget deficits more than doubled and the current account deficit fell to zero. This was
not an isolated incident. A similar story applies to the years 1980 to 1983.
Post Keynesian writing on fiscal policy has not all been defensive.
Lerners arguments of 1943 have been developed and qualified in important ways that relate to inflation, external balance and the public debt.
Much has been done on indicators of the stance of fiscal policy and the
mistake of trying to use fiscal policy to solve structural problems has been
pointed out. For a while after the Second World War it did seem possible
that there was a narrow zone of economic activity compatible with both full
employment and a very low rate of inflation. However, as more and more
workers were younger, with no memories of the depression of the 1930s, the
situation predicted by Michal- Kalecki emerged. Full employment reduced
substantially employers power to discipline workers, leading to declining
eciency and inflationary wage demands. Most Post Keynesian economists argued that, if fiscal policy was to be successful in maintaining the
economy at, or close to, full employment, it had to be supplemented with
an incomes policy. (See, for example, Cornwall 1983, chapters 11 and 12.)
As well as specific incentives or penalties in tax-based incomes policies (see
ibid., pp. 2725), fiscal policy can support incomes policies at the macro
level, for example, through a general trade-o between wage rises and tax
cuts or increased expenditure on the social wage. Experience with incomes
policies suggests that even successful ones are only eective for a limited
period of time, which can usually be measured in years rather than decades.
Ongoing innovation in designing incomes policies is important if fiscal
policy, together with other policies, is to maintain full employment without
inflation.
Inflation also interacts with balance of payment problems. Continuing
large budget deficits, especially if accompanied by large current account
deficits, may lead financial markets to fear an increase in the rate of inflation in a country and to withdraw financial investment, leading to a decline
in the value of the countrys currency on the foreign exchange market. The
resulting inflationary pressure can put stress on any incomes policy and
could lead to a depreciation/inflation vicious circle. Whether or not the concerns about budget deficits are well founded, the actions of financial
markets cannot be shrugged o. If more than one equilibrium position is
possible they may result in an economy reaching an equilibrium with a high
rate of unemployment. This is particularly the case where equilibrium is
path determined. While most Post Keynesians focus on the disequilibrium
152
Fiscal policy
path in the short to medium run, rather than some longer-term equilibrium
position, the conclusion is the same: namely that fiscal policy may have to
be modified to meet the fears of financial markets, for example, by ensuring that the budget is balanced or in surplus when the level of economic
activity is high. Financial market fears may be greater if a country has
persistent large current account deficits, and many would agree that this is
justified if the current account deficit is used to finance increasing consumption rather than increasing investment. Post Keynesian economists
have pointed out that this situation is a sign of structural imbalance and
structural change is necessary before fiscal policy can be eective in maintaining full employment.
If a countrys public debt is held by its own citizens, the liability (to taxpayers) is balanced by the assets of those who hold the debt. Nevertheless,
the consequences for income distribution may be important. In theory
these could be overcome through tax and other fiscal measures for redistribution. In practice, if the interest bill is large, this may not be feasible for
political and even administrative reasons. A large public debt relative to
GDP reduces the freedom of action with respect to fiscal policy and may
impose other burdens. A quasi sound finance argument, that the budget
should be balanced not over a year but over the business cycle, is too strict
as it ignores the eects of inflation and economic growth. If nominal gross
domestic product is growing there can be a positive budget deficit on
average over the business cycle without any upward trend in the ratio of
public debt to gross domestic product. Most Post Keynesians argue that a
deficit on average is usually necessary for the health of an economy (see, for
example, Bougrine 2000, various chapters).
Since Post Keynesian economists believe that fiscal policy is a major tool
in the very important task of managing aggregate demand, they are particularly interested in measuring the eects of fiscal policy. The discussion has
centred on the construction of a single-number indicator of the stance of
fiscal policy. Most would agree that single-number indicators are very inadequate given the varying multipliers that are attached to dierent categories of government expenditure and taxation. However, it seems impossible
to move media discussion and political arguments past a single-number
indicator, with the nominal budget deficit the most often used despite its
manifest flaws (see Eisner 1986). Attention has therefore focused on alternative measures of the budget deficit. It is well known that, while fiscal
policy aects the level of economic activity, the level of economic activity
aects the outcome of fiscal policy, with tax revenues falling during recessions and government expenditure on transfer payments automatically
increasing. An alternative measure of the stance of fiscal policy is the structural deficit, or the size the deficit would be at a benchmark level of high
Full employment 153

employment, but with the current expenditure and taxation laws. Most
commentators agree that the sale of public assets should be excluded when
calculating the structural deficit. Post Keynesians argue that the decline in
the value of the government debt (including currency) due to inflation
should be subtracted from the deficit. Eisner adds an additional point,
arguing that not only must one correct for inflation, but it is also necessary
to look at the market value of the public debt, not its face value. The former
fluctuates with changes in the interest rate. Most ignore this point, which
strictly speaking relates to monetary policy, not fiscal policy. Any calculation of structural deficits involves making a judgement about what level of
economic activity should be taken as a benchmark. While this may not
aect year to year changes in the structural deficit, it will certainly determine the size of this deficit and often will determine whether fiscal policy
is judged to be expansionary or contractionary. The selection of the benchmark is aected by the relative weight one gives to the dangers of inflation
and unemployment. Hence, no calculation of the structural deficit is completely objective.
J.W. N
See also:
Budget Deficits; Economic Policy; Full Employment; International Economics; Investment;
Monetary Policy; Taxation; Tax-based Incomes Policy.
References
Bougrine, Hassan (ed.) (2000), The Economics of Public Spending: Debts, Deficits and
Economic Performance, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Cornwall, John (1983), The Conditions for Economic Recovery: A Post-Keynesian Analysis,
Oxford: Martin Robinson.
Eisner, Robert (1986), How Real is the Federal Deficit?, New York: Free Press, Macmillan.
Lerner, Abba P. (1943), Functional finance and the federal debt, Social Research, 10 (1),
February, 3851.
Nevile, J.W. (2000), Can Keynesian policies stimulate growth in output and employment?, in
Stephen Bell (ed.), The Unemployment Crisis in Australia: Which Way Out?, Cambridge:
Cambridge University Press, pp. 14974.
Full Employment
Blessed are the extravagant, for theirs shall be full employment.
(Lekachman 1966, p. 94)
The term full employment can be traced back to William Pettys 1662
work, A Treatise on Taxes and Contributions, in which he argued that nonproductive labour could be supported as a consequence of the capacity of
producers of consumption goods to generate a surplus over and above their
154
Full employment
own subsistence. The classical economists did not consider full employment
specifically but J.-B. Say (1803) denied that a production economy could
ever suer a general glut which would otherwise have led to unemployment.
Says Law that supply creates demand became the epithet of classical and
neoclassical theory and its underlying reasoning still dominates orthodox
macroeconomics today. There was not a denial that unemployment could
occur but it was considered to be a manifestation of a temporary disruption, rather than being a generalized tendency of a capitalist production
system. A lack of consumption would become by definition an act of
investment. Moreover, it was argued that there was a strict separation
between output and price theory the so-called classical dichotomy. So
the existence of money posed no special problems. While J.C.L. Simonde
de Sismondi and Thomas Malthus demurred and argued that generalized
gluts could occur even if the savingsinvestment identity held, their analyses were flawed. It was Karl Marx, in his critique of Malthus, who provided
a modern Post Keynesian rationale for generalized gluts. Marx understood
that money could be held as a store of value and this behaviour interrupted
the sequence of sale and purchase. He also laid the foundations of multiplier theory by arguing, in Theories of Surplus Value, that, once this unity
of sale and purchase was disturbed, the chain of contractual relationships
between suppliers became threatened and overproduction, and then bankruptcies and unemployment, became widespread.
So, by 1900, there were two broad views about the possibility of full
employment: (a) Marxian views of crisis and the reserve army of unemployed, which saw capitalism as being incompatible with a fully employed
working class; and (b) the dominant (marginalist) view that unfettered
market operations would ensure that all those who wanted to work at the
equilibrium real wage could find it because Says Law held. Full employment became equivalent to the equilibrium intersection between the
demand for and supply of labour, which in turn reflected the productive
state of the economy driven by technology and the unconstrained preferences of the population. By definition, any workers who were idle were
voluntarily enjoying leisure and could not reasonably be considered unemployed. Mass unemployment was considered to be a transitory disturbance.
The advent of the Great Depression made it hard to justify the view that
the persistently high unemployment was due to changing preferences of
workers (increased quits in search of leisure), excessive real wages (in the
face of money-wage cuts), and/or a temporary interruption to market eciency. For the first time, notwithstanding Marxs inspiring insights,
Western economists articulated a macroeconomics that could define a
coherent concept of full employment and also explain mass unemployment
in terms of the inherent tendencies of monetary capitalism. The clue lay in
Full employment 155

recognizing the unique role that money could play in resolving the tensions
that uncertainty created in the decision-making calculus of decentralized
agents, but also in realizing that the fallacy of composition was endemic in
the prevailing (micro) explanations of unemployment.
Whether the 1930s marked the birth of Post Keynesian notions of full
employment is debatable. Post Keynesian theory has fractured origins, with
some practitioners seeing the labour market in Marxian, then Kaleckian
terms and others tracing their ancestry to Keynes and his General Theory.
Certainly, the attack against the marginalist faith in self-equilibration
mounted by Keynes (1936) and his monetary analysis was path-breaking. It
also more clearly outlined what we now mean by the term full employment.
Keynes linked full employment to national income levels, such that full
employment occurred at the level of output when all who want to work at the
going money-wage rates can find a job. Full employment was the absence
of involuntary unemployment. This was defined by the following thought
experiment: if a rise in nominal demand with constant money wages increased the price level (of wage-goods) but also resulted in both the demand
for and supply of labour increasing beyond the existing volume of employment then those who gained the new jobs were involuntary unemployed.
Involuntary unemployment was to be expected in a monetary economy
subject to uncertainty, because the act of holding money as a source of
liquidity provided the type of interruption to the outputspending balance
that Marx had clearly envisaged.
Consequently, the maintenance of full employment required government
policies to maintain levels of aggregate demand sucient to achieve output
levels consistent with all available labour being employed. Significantly, a
departure from full employment was construed as a systemic failure, rather
than an outcome related to the ascriptive characteristics of the unemployed
and/or the prevailing wage levels. Consistent with this notion was the coexistence of unfilled vacancies and unemployed workers as part of the normal
daily resolution of hiring and quits. Accordingly, full employment arose
when all unemployment was frictional. Beveridge (1944) defined full
employment as an excess of vacancies at living wages over unemployed
persons. The emphasis was on jobs.
Macroeconomic policy in the postwar period was designed to promote
full employment. Beveridge (1944, pp. 12335) argued that The ultimate
responsibility for seeing that outlay as a whole, taking public and private
outlay together, is sucient to set up a demand for all the labour seeking
employment, must be taken by the State. In the following years, a number
of Western governments, including those in Britain, Australia and Canada,
made a commitment to at least high and stable employment, if not full
employment. The US government was more circumspect, with its 1946
156
Full employment
Employment Act aiming only to ensure that employment opportunities

were maintained. From 1945 to the mid-1970s, most governments used
counter-cyclical budget deficits and appropriately designed monetary
policy to maintain levels of demand sucient to maintain full employment.
Unemployment rates in Western economies were at historical lows throughout this period.
In the 1950s, however, the emphasis on jobs was replaced by a concern
for inflation. Although only a subtle change, the redefinition of full employment in the early 1950s in terms of an irreducible minimum unemployment
rate (see Mitchell 2001) gave way soon after to the Phillips curve revolution.
The Keynesian orthodoxy considered real output (income) and employment as being demand determined in the short run, with price inflation
explained by a negatively sloped Phillips curve (in both the short run and
the long run). Policy makers believed they could manipulate demand and
exploit this trade-o to achieve socially optimal levels of unemployment
and inflation. The concept of full employment had been redefined to be the
rate of unemployment that was politically acceptable, given the accompanying inflation rate.
Milton Friedmans 1968 American Economic Association address and
the supporting work from Phelps (1967) provided the basis for the expectations-augmented Phillips curve, which spearheaded the resurgence of
pre-Keynesian macroeconomic thinking in the form of monetarism. Underpinning the natural rate hypothesis (NRH) was a unique cyclically-invariant
natural rate of unemployment (NRU), which was consistent with stable
inflation. There was no long-run, stable trade-o between inflation and
unemployment. The concept was broadened in the 1970s to incorporate a
number of structural labour market impediments, and the term non-accelerating inflation rate of unemployment (NAIRU) become popular.
The acceptance of these new ideas was aided by the empirical instability of the Phillips curve in most OECD (Organization for Economic
Cooperation and Development) economies in the 1970s following the
OPEC (Organization for Petroleum Exporting Countries) price rises.
Unemployment was considered to be voluntary and the outcome of optimizing choices by individuals between income and leisure. Full employment was assumed to prevail (with unemployment at the natural rate),
given the operation of market forces, unless there were errors in interpreting price signals. The NAIRU was now viewed as synonymous with full
employment. There was no discretionary role for aggregate demand management; only microeconomic reform would cause the NRU to change.
Accordingly, the policy debate became increasingly concentrated on deregulation, privatization and reductions in the provisions of the welfare state,
while the monetarist fight inflation first strategies ensured that unemploy-
Full employment 157

ment persisted at high levels. The NAIRU proponents responded by
claiming that the steady-state unemployment rate must have risen due to
worsening structural impediments, although they failed empirically to substantiate their argument. The fact that quits were strongly pro-cyclical
undermined the NRH, but the orthodoxy managed to avoid the damaging
empirical evidence. Full employment as conceived by Beveridge had been
abandoned.
With Post Keynesian economics dependent on the use of aggregate
demand management as a means of attenuating the fluctuating spending
patterns of the private sector (in particular, investment), the NAIRU
approach to inflation control presented a fundamental quandary.
The earlier approach to improving the Phillips curve trade-o was to
complement demand management policy with incomes policy, the latter
being designed to batten down the supply (cost) side. Some Post Keynesians
(principally the Marxian strain) had inflation models based on incompatible real income claims by workers and capital that delivered analytical findings observationally equivalent to the NAIRU approach. They also saw a
role for incomes policy although, following Marx and Michal- Kalecki, they
did not think that full employment (in the Keynesian sense) and capitalism
were compatible.
The modern Post Keynesian approach to the NAIRU challenge is best
represented by the hysteresis and persistence literature, although some of
the developments in this regard are strictly Neo-Keynesian. Hysteresis or
path-dependence was traced to various cyclical adjustments that occurred
in the labour market, which could be reversed in a growing economy. So
while the steady-state unemployment rate rose after a long downturn,
aggregate demand expansions could bring it down again. Once again full
employment could be achieved at relatively low unemployment rates
without ever-accelerating inflation. More recent empirical work has cast
doubt on the robustness of the NAIRU story and provided strong support
for a hystereticasymmetric interpretation of the inflationunemployment
relationship (Mitchell 2001).
Despite these developments, Post Keynesians cannot agree on the way to
pursue full employment. The predominant view assumes that the economy
is still amenable to a broad Keynesian spending expansion. Some Post
Keynesians eschew this approach, arguing that it will be inflationary and/or
environmentally damaging. They observe that the economies which avoided
the plunge into high unemployment in the 1970s all maintained a sector that
provided an employer of the last resort capacity to redress the flux and
uncertainty of private sector spending. In most countries, throughout the
1950s and 1960s, the public sector played this role, which ceased when the
monetarists began attacking the public sector on (orthodox) eciency
158
Full employment
grounds. Accordingly, these Post Keynesians propose a job guarantee

(Mitchell 2001) or an employer of last resort (Wray 1998), where the public
sector maintains a constant fixed-wage job oer to anyone who cannot find
employment elsewhere.
These models present a serious challenge to the conventional wisdom
that budget deficits are unsustainable, which is accepted by many Post
Keynesians. Wray and Mitchell argue that the issue of government securities is designed to sustain the target interest rate, set by monetary policy,
rather than being required ex ante to finance expenditure. They argue that
deflationary fiscal and monetary policy which drives employees out of the
private sector into lower-paid job guarantee jobs provides an adequate
counter-inflation policy if required.
Other Post Keynesians, including Arestis and Sawyer (1998), disagree.
They point to the possibility of higher inflation and an unsustainable
balance of trade leading to a growing ratio of debt to GDP. However, these
are problems associated with the pursuit of full employment per se, and not
the policies adopted to achieve it. A higher current account deficit as a ratio
of GDP may promote a depreciation, which would reduce the overall real
incomes of residents, but it may also be the price that must paid for
increased employment opportunities. In addition, Arestis and Sawyer
remain uncertain about the relationship between interest rates and budget
deficits and note the potential adverse reaction of the financial markets to
fiscal expansion.
Over the past 30 years the NAIRU concept has obfuscated the debate
over the capacity of capitalist economies to achieve and maintain full
employment, as traditionally understood. Somewhat belatedly this debate
is now occurring, but it remains unresolved within the deeply divided Post
Keynesian literature.
W M
M W
See also:
Budget Deficits; Economic Policy; Eective Demand; Employment; Fiscal Policy; New
Classical Economics; Says Law; Tax-based Incomes Policy; Unemployment; Wages and
Labour Markets.
Bibliography
Arestis, P. and M. Sawyer (1998), Keynesian economic policies for the new millennium,
Beveridge, W. (1944), Full Employment in a Free Society, London: Allen & Unwin.
Macmillan.
Lekachman, R. (1966), The Age of Keynes, Harmondsworth: Penguin.
Mitchell, W.F. (2001), The Job Guarantee and inflation control, in E. Carlson and W.F.
Fundamentalist Keynesians
159
Mitchell (eds), Achieving Full Employment, Supplement to the Economic and Labour
Relations Review, 12, 1026.
Phelps, E.S. (1967), Phillips curves, expectations of inflation and optimal unemployment over
time, Economica, 34, 25481.
Say, J.-B. (1803), A Treatise on Political Economy, Library of Economics and Liberty.
Retrieved 30 June 2002 from the World Wide Web: http://www.econlib.org/library/Say/
sayT0.html.
Vickrey, W.S. (1993), Todays task for economists (full-employment policy), American
Wray, L.R. (1998), Understanding Modern Money: The Key to Full Employment, Cheltenham,
UK and Northampton, MA, USA: Edward Elgar.
The term fundamentalist Keynesians originates with Coddington (1976)
to describe those who have seen Keyness work as a frontal assault on the
whole reductionist programme (p. 1259). The fundamentalist Keynesians
are those radical/Post Keynesians who see Keyness General Theory as a
rejection not only of the theories and policy prescriptions of neoclassical
economics but also of its analytical (reductionist) methods. In particular,
the fundamentalist Keynesians interpret Keynes as emphasizing the importance of uncertainty in economic behaviour and, as a consequence, rejecting the usefulness of both the optimization calculus and equilibrium
analysis.
Coddington associated the fundamentalist strand of Keynesianism primarily with George Shackle and Joan Robinson. Both Shackle and
Robinson believed the essence of Keyness revolutionary contribution to be
his analysis of the eects of uncertainty on investment in chapter 12 of the
General Theory, a theme that Keynes highlighted in his subsequent (1937)
Quarterly Journal of Economics article. Both also argued that Keyness
analysis implies the need for a fundamental change in the analytical
methods employed by economists. Robinson (1979) summarized the
methodological argument very succinctly: As soon as the uncertainty of
the expectations that guide economic behaviour is admitted, equilibrium
drops out of the argument and history takes its place (p. 126). Shackle
(1967) considered Keynes to have invented a scheme of thought for dealing
with the eects of uncertainty on economic behaviour. Shackle called this
scheme of thought the kaleidic analysis of a development through time in
which one situation or event grows out of another (p. 151, emphasis in
original). More recently, Davidson (1991) has restated the methodological
argument as the rejection of the ergodic axiom that allows uncertainty to
be modelled as a well-defined probability distribution. From Davidsons
perspective, Keyness concept of uncertainty represents a non-ergodic
160
process in which economic behaviour consists of crucial decisions and

unique events such that individuals cannot specify a complete set of possible future outcomes and associated probabilities.
Coddington dismissed the fundamentalist Keynesians as purely nihilistic as regards the development and practical application of economic
theory. He argued that the fundamentalists give too much emphasis to
those parts of Keyness thought that are merely a ground-clearing exercise.
For Coddington, fundamentalist Keynesians are on the slippery slope
towards subjectivism, driving a wedge between behaviour and circumstance that negates analysis and renders economic theorizing impossible.
There is some validity in the criticism of early fundamentalist Keynesians
as rather nihilistic, emphasizing mainly the negative aspect of Keyness
work. But Coddington himself was guilty of an excessive emphasis on the
negative aspect of Keynesian fundamentalism. He ignored, for example,
Shackles attempts to construct an alternative theory of decision making
under uncertainty using the concept of potential surprise. Coddington also
wrongly implied that Keyness concern with uncertainty was a transitional
phase. Keynes had a long-standing interest in probability theory dating
back to his undergraduate days. He published A Treatise on Probability in
1921 and explicitly referred back to this work in his discussion of uncertainty in chapter 12 of the General Theory.
The recognition of the centrality of probability and uncertainty throughout Keyness thought led to the emergence of a new Keynesian fundamentalism that sought to ground Keyness later economic analysis in his early
philosophical thought, especially A Treatise on Probability. This new
Keynesian fundamentalism originated with Lawson (1985) and the subsequent books by Carabelli (1988), Fitzgibbons (1988) and ODonnell (1989).
A central theme of the new fundamentalists is the relationship between
Keyness analysis of uncertainty in the General Theory and the logical
theory of probability that Keynes developed in A Treatise on Probability.
Although the emphasis of the new fundamentalists is on the task of interpretation, there is a clear implication that Keyness logical theory of probability may provide the basis for the development of an alternative
economic theory of behaviour under uncertainty.
Keyness logical theory of probability was an attempt to generalize
beyond the frequency theory of probability. Keynes defined probability as
the rational degree of belief in a proposition given the available evidence.
A probability is a rational degree of belief in the sense of being objectively
derived by logic rather than a matter of individual subjective evaluation.
Keyness concept of probability is epistemic in the sense of pertaining to
the nature of knowledge. In contrast, frequency theory treats probability as
an aleatory concept relating to the nature of the world. For Keynes, rela-
161
tive frequencies are a special type of quantitative data from which a numerical degree of belief in a proposition can be derived. Indeed, Keynes argued
that numerical probabilities are a special case that had been overemphasized because of their amenability to mathematical manipulation. Keynes
considered probabilities to be typically non-numerical and, in some cases,
non-comparable.
As well as the notion of probability as a rational degree of belief, Keynes
also introduced the concept of the weight of an argument. The weight of
an argument is a measure of the amount of evidence on which a proposition is based. For the most part, Keynes considered the weight of an argument to be the amount of relevant evidence. The weight of an argument is
independent of its probability. As additional relevant evidence is acquired,
the weight of an argument increases but the rational degree of belief in the
proposition may increase, decrease or remain unchanged. However, Keynes
is not entirely consistent in his definition of the weight of an argument. He
also referred to the weight of argument as the degree of completeness of
evidence, as well as the balance of absolute amounts of relevant knowledge
and relevant ignorance. These two alternative definitions of the weight of
argument imply the possibility that additional relevant evidence may
reduce weight if the assessment of relative ignorance is revised upwards.
Despite the conceptual diculties in formalizing the definition of the
weight of an argument, Keynes stressed that weight as well as probability
is relevant to practical decision making.
Keyness emphasis on the importance of the weight of an argument is a
key element in his critique of the doctrine of mathematical expectation as
a theory of human behaviour under uncertainty. The doctrine of mathematical expectation implies that alternative courses of action are evaluated
by weighting the value of the outcome with its probability. Keynes considered this approach to be too limited as a theory of human behaviour under
uncertainty. He argued that any such theory must incorporate not only the
value of the outcome and its probability as determinants of human
behaviour, but also the weight of the available evidence and the risk
attached (that is, the possible losses associated with any course of action).
Keynes considered the possibility of amending the doctrine of mathematical expectation by weighting the value of the outcome with what he termed
the conventional coecient instead of the probability. The conventional
coecient depends not only on the probability of the outcome but also on
weight and risk. The conventional coecient would tend towards the probability as weight tends towards unity and risk tends towards zero. However,
Keynes concluded that the conventional coecient is too restrictive in its
formalization of the eects of probability, weight and risk on uncertain
choices between alternative courses of action.
162
In attempting to interpret Keyness later economic writings as grounded in

his earlier philosophical thought, a crucial issue is the extent to which his philosophical position remained unchanged in any fundamental way between A
Treatise on Probability and the General Theory. The continuity-or-change?
debate has focused on four principal pieces of textual evidence: a biographical essay on F.Y. Edgeworth published in 1926; a letter to F.M. Urban, the
German translator of the Treatise, also written in 1926; a review of Frank
Ramseys Foundation of Mathematics in 1931; and Keyness autobiographical essay, My Early Beliefs that was read to the Bloomsbury Group in 1938.
These texts are ambiguous and subject to radically dierent interpretations.
There remains considerable controversy regarding whether Keynes retained
both the logical theory of probability and the belief that the inductive
method is only applicable to atomistic (as opposed to organicist) systems.
Disputes over interpretation such as these cannot be resolved in any definitive way by textual evidence, since ultimately they represent dierences over
the appropriate frame of reference. The only closure possible to the question
what did Keynes really mean? is to recognize the possibility of multiple
answers and to move on to another question. From the perspective of constructing an alternative non-neoclassical theory of economic behaviour
under uncertainty, the relevant question is whether or not Keyness analysis
in the General Theory can be developed by reference back to A Treatise on
Probability. In this case the emphasis shifts to the possibility of theoretical
continuity rather than its actuality in Keyness own intellectual development.
Keyness analysis of economic behaviour under uncertainty is fundamental to the logic of the General Theory. Keynes rejected the two foundations of neoclassical theory, namely, the aggregate labour market and Says
Law that supply creates its own demand, the latter justified theoretically by
the loanable funds theory of the rate of interest as the equilibrating mechanism ensuring that planned savings and investment are equated. Keynes
proposed the principle of eective demand in which the level of income
(and, in turn, output and employment) would adjust via the multiplier
process to bring savings automatically into line with the volume of investment. Hence, ultimately, the level of employment depends on the determinants of the volume of investment, ceteris paribus. Keynes argued that the
investment decision depends on prospective monetary yields (that is, the
marginal eciency of capital) exceeding the rate of interest. But prospective yields on investment depend on the state of long-term expectations.
There is, therefore, an essential link between uncertainty and involuntary
unemployment. If business is highly uncertain about future investment
prospects, the volume of employment will fall, leading to downward multiplier eects on income, output and employment. It is this essential link that
is emphasized by the fundamentalist Keynesians.
163
Keynes drew a crucial distinction in the General Theory between shortand long-term expectations. Short-term expectations relate to day-to-day
production decisions. These expectations are subject to continual revision
in the light of market outcomes. Mistaken short-term expectations can
cause temporary departures from the full-employment equilibrium, as had
been recognized by neoclassical economists prior to Keynes. In particular,
underestimation of current market demand is one of the causes of frictional unemployment. Mistaken short-term expectations do not cause
involuntary unemployment.
In contrast, long-term expectations relate to the estimation of the prospective monetary yields from investment projects. Keynes conceived of the
state of long-term expectations as consisting of two components: the most
probable forecast and the state of confidence. The latter refers to the degree
of uncertainty attached to the most probable forecast. It is at this point that
Keynes explicitly referred back to A Treatise on Probability and the concept
of the weight of an argument to clarify the meaning of uncertainty. It is
also consistent with his earlier criticism of the doctrine of mathematical
expectation as too limited. The investment decision depends not only on
the probability of alternative outcomes but also on the degree of confidence attached to these probability estimates, based on an assessment of
the amount of relevant evidence. This insight provides the basis for the construction of an alternative theory of economic behaviour under uncertainty. Keynes argued further that business recognizes the precariousness
of its estimates of prospective yields and, as a consequence, investment
decisions are not based purely on mathematical calculations. He recognized
that there are crucial non-rational elements in investment behaviour,
namely, an innate urge to action over inaction (that is, animal spirits) as well
as falling back on the conventional belief that the existing state of aairs
will continue unless there are specific reasons to expect particular changes.
Keyness analysis of economic behaviour under uncertainty in the
General Theory required a change in the method of equilibrium analysis.
Keynes retained the notion of equilibrium in the general sense of a position of rest but rejected the specific neoclassical definition of equilibrium
as a market-clearing allocative outcome. Keynes set out a three-stage
shifting equilibrium analysis (see Gerrard 1997). The first stage is to
determine the point of long-period equilibrium given a particular state of
long-term expectations. The second stage is the logical-time analysis of the
process of transition from one long-period equilibrium to another consequent on a shift in the state of long-term expectations. The final stage is
the dynamic analysis of historical time consisting of a complex of overlapping transitional processes arising from a multitude of changes in longterm expectations.
164
To conclude: fundamental Keynesians have been critical, and rightly so,

of neoclassical methods of analysis. To the extent that this criticism is
grounded on Keyness detailed analysis of probability, long-term expectations and equilibrium, the fundamentalists are not nihilistic but rather
should be seen as providing the foundations for the construction of a truly
radical Keynesian alternative.
B G
See also:
Equilibrium and Non-equilibrium; Expectations; Keyness General Theory; Non-ergodicity;
Time in Economic Theory; Treatise on Probability; Uncertainty.
References
Carabelli, A. (1988), On Keyness Method, London: Macmillan.
Coddington, A. (1976), Keynesian economics: the search for first principles, Journal of
Economic Literature, 14 (1), 125873.
Davidson, P. (1991), Is probability theory relevant for uncertainty? A Post-Keynesian perspective, Journal of Economic Perspectives, 5 (1), 12944.
Fitzgibbons, A. (1988), Keyness Vision: A New Political Economy, Oxford: Clarendon Press.
Gerrard, B. (1997), Method and methodology in Keyness General Theory, in G.C. Harcourt
and P.A. Riach (eds), A Second Edition of The General Theory. Volume 2, London:
Routledge, pp. 166202.
Lawson, T. (1985), Uncertainty and economic analysis, Economic Journal, 95 (380), 90927.
ODonnell, R.M. (1989), Keynes: Philosophy, Economics and Politics, London: Macmillan.
Robinson, J. (1979), History versus equilibrium, in J. Robinson, Contributions to Modern
Economics, Oxford: Basil Blackwell, pp. 12636. (First published in Thames Papers in
Political Economy, London, 1974.)
Shackle, G.L.S. (1967), The Years of High Theory: Invention and Tradition in Economic
Thought 19261939, Cambridge: Cambridge University Press.
Globalization
Globalization is not a term invented by economists, and thus it is one of
which all economists Post Keynesian and mainstream alike are suspicious, despite (or, in some cases, because of) the fact that it is at the centre
of many contemporary economic policy debates. To the extent that globalization is considered synonymous with liberalization, Post Keynesians have
been outspoken sceptics. In the face of the dominant neoliberal economic
model with its call for financial market deregulation and fiscal and monetary austerity, Post Keynesians have insisted instead on expansionary
macroeconomic policies and controls on international capital movements.
Globalization can be seen as a two-part process the globalization of
production and the globalization of finance. While both parts are the result
of heightened international capital mobility, the globalization of finance is
understood through the Post Keynesian theory of markets, while the analysis of the globalization of production requires the Post Keynesian theory
of the firm and oligopoly. The globalization of production comprises international trade and foreign direct investment, and while the Post Keynesian
theory is less well developed in these areas than in the area of finance, it
none the less provides the building blocks for a rich description and policyrelevant theory of globalized production. Below we consider each of the
aspects of globalization in turn.
Post Keynesians are generally sceptical of the global benefits of international financial market liberalization for two basic reasons. The first follows
from the general Post Keynesian view that market flexibility does not bring
optimality (for example, full employment) since the problem of unemployment is the result neither of market rigidities nor of information distortions
resulting from government intervention or imperfect competition. Failures
of eective demand can exist in the absence of either of these conditions.
Moreover, price movements alone (through wages or exchange rates) are
unlikely to bring about large adjustments in international payments imbalances and are swamped by the eect of changes in income and demand.
Accordingly, international dierences in the rate of economic growth and
thus international divergence of incomes are explained in Post Keynesian
theory by international dierences in the income propensities to export and
import.
The second is related to the risk of capital flight that comes with capital
market liberalization. Post Keynesians have relied on Keyness distinction
between speculation and enterprise, the former referring to the activity
165
166
Globalization
of forecasting the psychology of the market and the latter the activity of
forecasting the prospective yield of assets over their whole life (Keynes
1936 [1964], p. 158). Keynes noted that capital markets national or international can at times be dominated by speculative behaviour that can
move the economy away from full employment. In an oft-cited passage, he
wrote:
Speculators may do no harm as bubbles on a steady stream of enterprise. But
the position is serious when enterprise becomes the bubble on a whirlpool of
speculation. When the capital development of a country becomes a by-product
of the activities of a casino, the job is likely to be ill-done. (Keynes 1936 [1964],
p. 159)
Keynes saw the eects of speculation to be particularly detrimental in an

open economy context when there is a risk of capital flight. In his 1933
essay, National self-suciency, Keynes argued that the ability of the state
to pursue full employment (monetary and fiscal) policy may be jeopardized
by international capital mobility. Thus, Keynes wrote, [L]et goods be
homespun wherever it is reasonably and conveniently possible, and, above
all, let finance be primarily national (Keynes 1933 [1982], p. 236).
For these two reasons, Post Keynesians have been sceptical of flexible
exchange regimes and of capital market liberalization generally. Price
inelasticity of trade explains the ineciency of exchange rate adjustment,
while the volatility of liberalized capital markets gives support to the policy
of capital controls. Post Keynesians make the empirical argument that the
rapid rates of economic growth experienced during the era of Bretton
Woods resulted, in part, from the limits on the international mobility of
capital and the relative fixity of exchange rates. Post Keynesians have typically found the source of Asian economic crises of the 1990s in the excessive (or too rapid) liberalization of foreign capital markets and have
supported the use of bank-based rather than equity-based financing for
economic development on the grounds that the latter encourages excessive
speculation (and capital flight) rather than entrepreneurship.
Scepticism towards capital market liberalization has led to a variety of
proposals for the regulation of international capital flows, including a
transactions tax on international capital flows, an international reserve and
capital adequacy requirement on all financial corporations, international
procedures for the orderly sorting out of competing claims in the case of
default on sovereign debt, or the establishment of a new central bank clearing unit to promote expansionary payments adjustment rather than the
contraction that occurs in the current system.
The starting-point of the Post Keynesian theory of the globalization of
production is the recognition that, in a market economy, unemployment,
Globalization
167
excess capacity and international payments imbalances have no natural

tendency to reverse themselves. For a country operating at full employment, a payments imbalance can be expected to bring pressure for a change
in the price level, as posited by the price-specie-flow mechanism. In the
presence of persistent unemployment, trade imbalances will bring about
movements in the rate of interest rather than in the price level. Trade imbalance results not in a change in the price level but in a potential liquidity
problem for the deficit country by reducing the monetary base. This, in
turn, will alter the interest rate. A surplus on current account will, by
similar logic, reduce the rate of interest. Interest rate increases might, of
course, move the economy further from full employment. Keynes himself
argued that under certain conditions the balance of payments is the main
determinant of the rate of interest, in which case improving the balance of
payments is essential for the attainment of full employment.
Three important implications emerge for the understanding of globalization. First, without a well-functioning price-specie-flow mechanism, trade
imbalances may persist over long periods of time. The balanced trade
implication of the principle of comparative advantage is, in this way, equivalent to Says Law in an open economy, whereby export growth automatically generates an equivalent increase in imports, or vice versa. Second, the
direction of international trade, and thus the international division of
labour, will be determined by absolute rather than comparative advantage
since the mechanisms which would otherwise transform a situation of dierential comparative costs into one of dierences in absolute money costs
and prices no longer operate. That is, the adjustment is simply not adequate
to guarantee that the principle of comparative advantage will determine the
direction of trade and a zero payments balance for all countries (Milberg
2002). According to Robinson (1973, p. 16), The comforting doctrine that
a country cannot be undersold all round was derived from the postulate
of universal full employment. The argument consists merely in assuming
what it hopes to prove. Finally, if trade is determined by absolute advantage and countries can indeed be undersold all round, then free trade is
not necessarily the first-best policy, since infant industry protection may be
needed to spur technical change needed for international competition.
The other aspect of globalized production is foreign direct investment,
and Post Keynesian pricing theory provides some relatively untapped
insights. Hymer, building on Ronald Coases emphasis on transactions costs
and Alfred Chandlers focus on the historical evolution of corporate capitalism, was the first to understand that the phenomenon of foreign direct
investment was necessarily driven by oligopolistic firms. The high volume of
cross-hauling (that is, simultaneous inward and outward foreign direct
investment in one country) implies that the process is not driven simply by
168
Globalization
arbitrage of temporarily high profit opportunities in one location compared

to another. Hymer, and later others, argued that the transnational firm is a
non-market institution, and its desire to internalize international operations
constitutes a market failure, but is the prime reason for firms to invest
abroad rather than serve foreign markets in other ways, such as exports. In
oligopoly, firms are large and few, or as Hymer puts it, the size of the
market is limited by the size of the firm (Hymer 1970, p. 443).
The oligopoly corporation emerged in the late nineteenth century as the
organizational form that best captured economies of scale, best avoided the
otherwise destructiveness of price-based, perfect competition, and insulated investment from cyclical downturns. Transnational corporate investment began as oligopolies matured in the 1920s. Over time, foreign direct
investment became a new weapon in the arsenal of oligopolistic rivalry
(Hymer 1972, p. 444) as firms sought new markets, and the control of
resources and cheap labour all the while conserving their transactions
cost advantage over market-based operations such as through exports.
Post Keynesians have long recognized the ruinous nature of price competition and thus the necessity of oligopoly over the long run (Eichner
1976, p. 11). More important, such a recognition has led to an alternative
theory of price determination in capitalism, in which the firm, rather than
market forces of supply and demand, plays the dominant role. According
to Shapiro and Mott (1995, p. 38), The prices derived in the mark-up
models of the [Post Keynesian] theory are not the prices that serve the
unconscious ends of the market (the allocative eciency of the neoclassical theory or the systemic reproduction of the Ricardian conception) but
the ones that serve the conscious ends of the enterprise. In Eichners (1976,
chapter 2) theory of the megacorp, firms use pricing as a means to generate finance for future investment. From the perspective of the transnational
corporation, international investment allows the internalization not only of
firm-specific advantages related to technology, management, marketing
and so on, but also the internalization of the pricing decision on international (intra-firm) transactions.
Does the recent trend towards outsourcing and subcontracting constitute a reversal of the oligopolistic trend identified by Chandler, Hymer,
Eichner and others? The process has become so prevalent that the contemporary manufacturing firm often does no manufacturing at all. Most outsourcing relations today are arms length in a formal sense only. The rise
in outsourcing and subcontracting constitutes a sharpening of the hierarchical structure that Hymer identified with the modern corporation, due
to the added flexibility that outsourcing provides and the selective competition (among suppliers) that it promotes. Subcontracting is driven by the
desire of firms to increase flexibility and lower unit labour costs. Cost
Globalization
169
reduction can come with increased productivity or lower wages. In this

sense, the rise of international outsourcing reintroduces the ruinous competition from which capitalism escaped in the late 1800s. The expansion of
sweatshop labour is thus an integral part of the globalization of production,
and is another source of rising income inequality in developing countries.
The insights of Hymer and Eichner can form the foundation of a Post
Keynesian theory of international production, but the full theory remains
undeveloped and untested. Moreover, a truly Post Keynesian theory will
make a connection between the finance and production processes in the
global economy. This becomes especially important as firms outsource production operations and focus increasingly on financial management.
Investment location decisions may themselves be influenced by foreign
exchange portfolio considerations, for example. That is, the manufacturing
firm is increasingly a financial unit. Keyness distinction between speculation and enterprise is being blurred further as foreign direct investment can
increasingly be hedged with the build-up of domestic liabilities.
In the conclusion of The General Theory, Keynes wote that The outstanding faults of the economic society in which we live are its failure to
provide for full employment and its arbitrary and inequitable distribution
of wealth and incomes (Keynes 1964, p. 372). These faults are arguably
worse today than they were 25 years ago, when the current wave of globalization began. The global economy suers chronic unemployment, excess
capacity in most manufacturing sectors and growing income inequality.
Real wage suppression and lax social standards in poor countries has not
brought them a degree of international competitiveness sucient to generate economic development; financial liberalization has hastened economic crises in East Asia, Russia and Central and South America; and
austerity imposed by the International Monetary Fund has tended to
worsen these problems. The economic logic of a policy of sustained global
demand expansion with regulated international capital mobility is reasonably well established, but the political obstacles to its implementation
remain large.
W M
See also:
Bretton Woods; Competition; Economic Policy; Exchange Rates; International Economics;
Pricing and Prices; Tobin Tax.
References
Eichner, A.S. (1976), The Megacorp and Oligopoly, Cambridge: Cambridge University Press.
Hymer, S. (1970), The eciency (contradictions) of multinational corporations, American
Hymer, S. (1972), The multinational corporation and the law of uneven development, in J.
170
Growth and income distribution
Bhagwati (ed.), Economics and the World Order: From the 1970s to the 1990s, New York:
Free Press, pp. 43663.
Keynes, J.M. (1933 [1982]), National self-suciency, Yale Review, Summer. Reprinted in The
Collected Writings of John Maynard Keynes. Volume XXI: Activities 19311939, London:
Macmillan for the Royal Economic Society, pp. 23346.
Keynes, J.M. (1936 [1964]), The General Theory of Employment, Interest and Money, New
York: Harcourt, Brace & Jovanovich.
Milberg, W. (2002), Says Law in the open economy: Keyness rejection of the theory of comparative advantage, in S. Dow and J. Hillard (eds), Keynes, Uncertainty and the Global
Economy, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 23953.
Robinson, J. (1973), The need for a reconsideration of the theory of international trade, in
R. Swoboda (ed.), Current Issues in International Economics, London: Macmillan, pp.
1525.
Shapiro, N. and T. Mott (1995), Firm-determined prices: the Post-Keynesian conception, in
Paul Wells (ed.), Post-Keynesian Economic Theory, Boston: Kluwer Academic Publishers,
pp. 3548.
Growth and Income Distribution

The determinants of growth were a major concern of the classical economists, who related growth to income distribution. The latter aects the
saving decisions of the dierent classes and, according to some classical
economists, their investment decisions too. Adam Smith and Karl Marx
also underlined the role of technical progress, presenting a broad analysis
of this subject, which can be considered an antecedent of the modern
cumulative causation and evolutionary approaches.
The rise of the neoclassical school in the second half of the nineteenth
century brought about a change of perspective in economic theory.
Allocation of resources became the major concern and the problem of distribution was seen as one aspect of the general pricing and allocation
process. Neoclassical economists argued that competitive forces, operating
through variations in relative prices and factor substitution, generate a tendency to full employment and to the exploitation of the growth potential
of the economy. These market mechanisms were examined in what Keynes
called the real department of economics. The monetary department
dealt instead with business fluctuations, arguing that the working of the
credit system cause or amplify them.
The severity of the Great Depression changed the course of these events.
As Roy Harrod pointed out (see Young 1989, pp. 3038), previous recessions had not led the economy too far from full employment, nor had they
cast doubt on the belief that the economy is able to return to it. The Great
Depression, however, endangered political stability and raised the problem
of a new political approach and of a new economic theory able to clarify
whether market forces can lead the economy towards full employment or
government intervention is required to restore it.
Growth and income distribution 171

Moving along these lines, in 1932 Keynes introduced the concept of a
monetary theory of production to attack the neoclassical separation
between the real and the monetary departments of economics and the idea
of a tendency to full employment. Harrod, on the other hand, began in
1933 to develop economic dynamics. His work was stimulated by the will
to extend Keyness ideas to the dynamic context. The seminal An essay in
dynamic theory thus conceived modern growth theory as a Keynesian
theory: it developed the views that the economic system does not tend necessarily to full employment and that the rate of growth may be aected by
the autonomous components of aggregate demand, coming from the
government, the private and the foreign sectors.
In opposition to Harrods views, Robert Solow presented in 1956 a
dynamic version of neoclassical theory. He argued that variations in relative
prices and factor substitution led the economy to a full-employment steady
growth path. The debate on capital theory, enhanced by the publication in
1960 of Piero Sraas Production of Commodities by Means of Commodities,
scrutinized Solows conclusions. Some outstanding neoclassical economists
acknowledged the validity of some criticisms raised against their theory.
Paul Samuelson recognized, in the summing up of the 1966 Symposium in
the Quarterly Journal of Economics, that in the long-period analysis of an
economy where more than one commodity is produced, the occurrence of
reverse capital deepening is the general case. This conclusion undercut
the neoclassical parables that extended to a multi-commodity economy the
conclusions from the analysis of a one-commodity world and challenged the
view that price variations and factor substitution lead the economy to full
employment.
During the same years, Kaldor (195556) and Pasinetti (1962) developed
the Post Keynesian theory of growth and distribution by assuming that
market forces operate along lines that are dierent from those envisaged by
neoclassical authors and similar to those described by the classical economists. Like the latter, Kaldor and Pasinetti assumed that the propensities to
save of dierent income earners (or classes, or sectors of the economy) are
not equal, and argued that variations in income distribution bring about
variations in total saving and aggregate demand, leading the economy to
steady growth. The Post Keynesian theory of growth and distribution introduced the Cambridge equation and the Pasinetti theorem, which state
that in steady growth the rate of profit is equal to the ratio between the rate
of growth and the capitalists propensity to save, and does not depend on
technology or on the workers propensity to save. In 1966 Samuelson and
Franco Modigliani challenged this conclusion and proposed an antiPasinetti or dual theorem. They argued that in steady growth, if the capital
owned by the capitalist class is zero, the capitaloutput ratio is equal to the
172
ratio between the workers propensity to save and the rate of growth, while
the rate of profit depends on the technological relation connecting this variable to the capitaloutput ratio. Whether the Pasinetti or the dual
theorem applies depends on this technological relationship too.
By focusing on the role of income distribution in the growth process, and
underlining the links with the classical economists and the dierences with
neoclassical authors, the theory proposed by Kaldor and Pasinetti failed to
emphasize that there is no automatic tendency to full employment.
However, developments in the KaldorPasinetti theory, which have examined the role of the demand coming from the government sector, atone for
this failure.
Kaldors 1958 Memorandum to the Radclie Commission shows many
similarities with the views on the role of government policy proposed by
Harrod and other Keynesian authors. Kaldor considered government policies to be necessary to pursue stability and growth. For him, monetary
policy is the appropriate tool against economic fluctuations, while fiscal
policy is relevant to the long-range objective of sustained growth. He proposed to use the Cambridge equation to determine the tax rate compatible
with both the full-employment rate of growth and the rate of interest fixed
by the monetary authority. In doing so he showed awareness of the complexity of the growth process, when he anticipated the view, developed
some years later, that an expansionary fiscal policy may cause problems for
the international competitiveness of the economy and for the maintenance
of sustained growth in the future.
Kaldor did not present his conclusions on the role of government policy
in a formalized way. The first formal presentation of the Post Keynesian
theory of growth and distribution, which explicitly introduces the government sector, was provided by Ian Steedman in 1972. He assumed a balanced
budget to show that the Cambridge equation in a revised form that takes
into account the existence of taxation holds in a larger number of cases
than the dual theorem of Samuelson and Modigliani. By the late 1980s,
Fleck and Domenghino and Pasinetti started a debate on the validity of the
Cambridge equation when the budget is not balanced. The debate examined
a large number of cases, showing when the Cambridge equation holds and
confirming the conclusions previously reached by Steedman (see Panico
1997). Its results describe how the views presented by Kaldor to the
Radclie Commission can be formally developed, clarifying some features
of his proposals. Moreover, they show the existence of some other common
elements between the classical and the Post Keynesian traditions, allowing
the reconciliation of two approaches to distribution, which had previously
been considered alternative. These are the approach proposed by Kaldor
and Pasinetti in their theory of growth and distribution and that implied by
Growth and income distribution 173

Sraas hint in Production of Commodities to take the rate of profit, rather
than the wage rate, as the independent variable (determined, in turn, by the
money interest rates) in the classical theory of prices and distribution.
Another line of development in the Keynesian literature on growth
focuses on the demand coming from the private sector in the form of autonomous investment. This literature presents several investment-driven growth
theories based on dierent specifications of the investment function.
The neo-Keynesian theory, proposed by Joan Robinson and Kaldor,
assumes a direct functional relationship between investment and the rate of
profit. This theory, which determines growth and distribution simultaneously, extends to long-period analysis the paradox of thrift, according to
which an increase in the propensity to save causes a reduction in the rate of
profit and in the rate of growth. Moreover, it underlines the existence of an
inverse relationship between the real wage rate and the rate of growth.
The Kaleckian theory, inspired by Michal- Kalecki and Josef Steindl,
assumes that (i) productive capacity is not utilized at its normal level, (ii)
the profit margin is an exogenous variable depending on the degree of
monopoly enjoyed by oligopolistic firms, (iii) prices are determined
through a mark-up procedure, and (iv) investment is positively related to
the rate of profit, which is a proxy for the state of expected profitability and
the availability of internal finance, and the degree of capacity utilization,
which reflects the state of aggregate demand. This theory confirms the neoKeynesian conclusion on the paradox of thrift and argues, in opposition to
the neo-Keynesian theory, for the existence of a positive relationship
between the real wage rate and the rate of growth in the presence of longrun underutilization of capacity. This result, known as the paradox of
costs, is due to the fact that the rise in the real wage rate brings about an
increase in demand and capacity utilization, which has a positive eect on
the rate of profit and on investment.
The Kaleckian theory has been recently amended by work inspired by
Bhaduri and Marglin (1990), which takes into account the dierent eects
on investment of the rate of profit, the profit margin and capacity utilization. By introducing an investment function positively related to the profit
margin and to capacity utilization, these works identify a wage-led and a
profit-led growth regime. In both cases, a rise in the real wage rate reduces
the profit margin and increases capacity utilization. However, in the wageled regime the overall eect of an increase in the real wage rate on growth
is positive, as in the Kaleckian paradox of costs, because the positive eect
on growth generated by the increase in capacity utilization is assumed to be
greater than the negative eect on growth generated by the decrease in the
profit margin. In the profit-led regime the opposite result holds, because the
positive eect on growth generated by the increase in capacity utilization is
174
assumed to be lower than the negative eect generated by the decrease in

the profit margin.
Finally, an attempt has been made in recent literature to develop a neoRicardian theory of growth, which starts from a classical theory of prices
and distribution. This theory, in opposition to the neo-Keynesian and
Kaleckian theories, assumes that the investment function depends on the
discrepancies between actual and normal capacity utilization and underlines the need to develop the analysis of growth through the comparison
of long-period positions. Moreover, it makes the rate of profit depend on
the money rate of interest, as suggested by Sraa in Production of
Commodities.
The last line of development of Keynesian literature focuses on the influence on growth of demand coming from the foreign sector, a problem
already considered by Harrod in the 1930s. This literature plays down the
role of distributive variables and is intertwined with the analysis of growth
as a cumulative process.
In a series of essays written between 1966 and 1972, Kaldor used the
notion of cumulative causation to describe the actual performance of
economies. He attributed to the demand coming from the foreign sector the
primary role in setting in motion the growth process. The domestic sources
of demand mainly influence, instead, the competitiveness of the economy
and the intensity with which an external stimulus is transmitted to the rate
of growth. According to Kaldor, the composition of output and demand
has an important influence on the rate of change of productivity, owing to
the presence of variable returns in the dierent sectors of the economy and
to the fact that increasing returns occur mainly in the capital-goods sector.
For Kaldor, high ratios of investment to aggregate demand and of the
capital-goods sector in the productive structure enhance productivity
changes, which, in turn, improve the international performance of the
economy, setting up and intensifying cumulative processes. He distinguished between the concepts of consumption-led and export-led
growth. The latter, he argued, is more desirable than the former, which
tends to have negative long-run eects on productivity and international
competitiveness, since it increases the weight of non-increasing returns
sectors in the productive structure of the economy. This distinction was at
the basis of Kaldors claim, noted above, that the maintenance of sustained
growth in the future may be endangered by the use of fiscal policy, which,
according to him, tends to increase the share of consumption in aggregate
demand.
In 1975, Robert Dixon and Anthony Thirlwall presented an export-led
growth model, which formalized some aspects of Kaldors views. Thirlwall
(1979), on the other hand, worked out a dynamic analysis showing how
Growth theory 175

growth may be constrained by the equilibrium of the balance of payments,
disregarding the operation of cumulative processes. In spite of this simplification, the empirical applications of the new analysis, which are able to
account for dierences in the rates of growth among countries and the
cumulative divergence in their GDP levels, have produced more satisfactory
results than those of the 1975 export-led model. Recently Moreno Brid
(199899) and McCombie and Thirlwall (1999) have extended Thirlwalls
new analysis to take into account the impact of the persistent accumulation of external debt on the economys long-term rate of expansion. These
extensions have opened new areas of research into the financial restrictions
imposed by international credit institutions on the long-term economic
growth of countries with persistent trade balance deficits.
C P
See also:
Cambridge Economic Tradition; Capital Theory; Growth Theory; Income Distribution;
Investment; Kaldorian Economics; Kaleckian Economics; Sraan Economics.
References
Bhaduri, A. and S. Marglin (1990), Unemployment and the real wage: the economic basis for
contesting political ideologies, Cambridge Journal of Economics, 14 (4), 37593.
Kaldor, N. (195556), Alternative theories of distribution, Review of Economic Studies, 23
(2), 83100.
McCombie, J. and A.P. Thirlwall (1999), Growth in an international context: a post
Keynesian view, in J. Deprez and J.T. Harvey (eds), Foundations of International
Economics: Post Keynesian Perspectives, London: Routledge, pp. 3590.
Moreno Brid, J.C. (199899), On capital flows and the balance-of-payments constrained
growth model, Journal of Post Keynesian Economics, 21 (2), 28398.
Panico, C. (1997), Government deficits in the Post Keynesian theories of growth and distribution, Contributions to Political Economy, 16, 6186.
Pasinetti, L.L. (1962), Rate of profit and income distribution in relation to the rate of economic growth, Review of Economic Studies, 29 (4), 10320.
Thirlwall, A.P. (1979), The balance of payments constraint as an explanation of international
growth rate dierences, Banca Nazionale del Lavoro Quarterly Review, 128, pp. 4553
Young, W. (1989), Harrod and His Trade Cycle Group, London: Macmillan.
Growth Theory
Sustained but irregular and unevenly distributed growth in output has been
a defining aspect of capitalism. An equally defining aspect of Post
Keynesian economic analysis has been its desire to realistically confront
this complex phenomenon. The undeniable diculty of this task in part
explains why the Post Keynesian literature on growth is as irregular and
uneven as its subject matter.
Post Keynesian growth models range from those that focus solely on the
176
Growth theory
phenomenon of growth and eschew the analysis of cycles, to others in

which cycles and growth are inseparable. However, most if not all Post
Keynesian growth models are distinguishable from neoclassical models by
considering at least two of the factors of growth, cyclicality and distribution (where the last factor concerns social classes, or dierent sectors of
industry, or both). There is therefore a strong overlap between Post
Keynesian treatments of growth, cycles, distribution and industrial structure. All are intrinsically entwined in Post Keynesian analysis and the real
world.
Harrod set the tone for subsequent work by Post Keynesians when, in
criticizing the neoclassical proclivity to treat growth and cycles as independent phenomena, he stated that the trend of growth may itself generate
forces making for oscillation (Harrod 1939, pp. 1415). His single-sector
model of unstable growth was driven by a savings function that depended
upon the level of output, and an investment function that depended upon
the rate of change of output. From this he derived a formula that equated
the savings to output ratio to the product of the rate of growth, and the
ratio of investment to change in output (the latter is known as the incremental capital to output ratio, or ICOR). This equality applied both to the
actual recorded rate of growth and ICOR, and the desired rate of growth
and desired ICOR that together fulfilled capitalist expectations. If the
actual rate of growth exceeded the desired rate, then the actual level of
investment would be below the desired level which then led to an increase
in investment that accelerated the rate of growth even further; the reverse
mechanism applied if actual growth was below the desired rate.
Joan Robinson, Nicholas Kaldor, Luigi Pasinetti, Richard Goodwin and
Edward Nell set o dierent analytic streams subsequent to Harrods
seminal contribution. Each stream involved some compromise forced by
the diculty of modelling the dynamic process of growth, though none
compromised realism as completely as was commonplace in neoclassical
theory. Today a substantial new band of nonlinear modellers are slowly
blending these historic roots with modern nonlinear mathematical methods
and computational analysis.
Robinson overcame the pre-computer inability to model growth processes out of equilibrium with the mental device of comparative golden
age economies. The structure of an economy was clearly specified in terms
of classes (workers, capitalists, rentiers), sources of income (wages, profits,
rent/interest), expenditures (consumption, investment, placement), industry sectors (consumption, investment), and fundamental rates of change
(population growth, technical progress). The proportions between these
variables that would be needed to ensure the highly unlikely outcome of
stable growth were then worked out, and two economies were assumed to
Growth theory 177

be in this golden age (sometimes with diering key values, such as the level
of real wages). A change in behaviour could then be postulated in one
economy (for example, an increase in birth rates leading to a rise in unemployment, or an increase in technical progress in investment goods) that
would move it o its golden path, and the change in systemic behaviour was
evaluated with respect to the economy that continued in its golden state.
Kregel (1975) gives a very accessible overview of Robinsons method, and
provides a useful survey of the rival approaches of Kaldor and Pasinetti.
Kaldor extended Harrods model by incorporating the topic of income
distribution between workers and capitalists, where capitalists had a higher
propensity to save than workers. Using the extreme assumption that
workers do not save and capitalists do not consume, he linked the rate of
profit to the rate of growth. Kaldor eschewed the concept of an aggregate
measure of capital, and argued that since technical progress was embodied
in new machines, capital in use would have a profile from the most profitable new machinery to the near-obsolete that would earn a zero rate of
profit.
However, while Robinsons approach emphasized the extreme improbability of any economy ever being on a golden path, and Harrods model had
an unstable equilibrium, Kaldor made the opposing assumptions that longrun growth had to involve the full employment of labour, and that the longrun equilibrium was stable. His reason for these assumptions that growth
concerns long-period analysis and only a full-employment equilibrium could
prevail in the long term would not be accepted today, since it is well known
that models of complex systems do not have to converge to an equilibrium
but can remain indefinitely in a far-from-equilibrium state. This assumption also drove a wedge between Kaldors short-run Keynesianism and his
long-run analysis.
Pasinetti corrected Kaldors model to allow for workers owning a proportion of profits, but concluded that the rate of growth was nevertheless
determined by the accumulation decisions of capitalists alone.
Kaldor also contributed a weather vane to economic analysis by
arguing that there were a number of stylized facts that any theory of
growth had to explain if it were to be regarded as prima facie tenable. These
included the primacy of the rate of growth of the manufacturing sector in
determining overall growth via the technological progress and increasing
returns to scale that emanate from this growth, the decline in agricultural
employment over time, and the relative constancy of income shares over
time. Subsequent Post Keynesians have added eective demand growth as
a key constraint on overall growth, a secular decline in manufacturing and
rise of service employment in advanced economies, and the need for models
of growth in which the monetary and financial system plays a crucial role.
178
Growth theory
Goodwins growth cycle, published in the centenary year of Marxs

Capital, gave a mathematical rendition of Marxs chapter 25 verbal model
of a growth cycle driven by a struggle over the distribution of income
between workers and capitalists in which the rate of unemployment acted
as a check upon workers wage demands. Technically the model was a
descendant of the predatorprey model first developed in biology by
Alfred Lotka and Vito Volterra, while economically it included the complete panoply of growth, cycles and income distribution. Although criticized in some quarters for structural instability and a lack of empirical
verification, it has been used as a basis for many more general models and
is still a fruitful basis for further research. Blatt (1983) provides a clearer
explanation of this model than can be found in Goodwins original writings.
Kaldor aside, most Post Keynesian work on growth has presumed that
the main constraint on the rate of economic growth comes not from supplyside issues as in neoclassical theory, but from eective demand constraints
and the non-neutrality of money.
Post Keynesians have also been interested in explaining why growth
occurs, as well as finding means to model it.
Kornai (1990) argues that firms in capitalist economies are constrained
not by resources and productivity, as in the neoclassical model, but by
limits to eective demand. This demand constraint means that firms
operate with significant excess capacity, since without this they can neither
respond to changes in the structure of demand, nor take advantage of
problems that might beset competitors. As a result, production costs do not
vary with output, and firms compete by product innovation rather than via
price: price competition is the exception rather than the norm.
Product innovation in turn requires research, development and investment, which both generates growth and gives rise to waves of Schumpeterian
creative destruction that give growth in capitalist economies its cyclical
nature. This analysis thus grounds technical progress and growth in productive capacity in the competitive interactions of firms vying for profit and
market share.
Nell (1998) emphasizes the transformational nature of growth in a capitalist economy growth involves not merely quantitative increase, but also
qualitative change in the composition of output, the nature of economic
institutions, and the state of economic expectations. Like Kornai, Nell
stresses the role of real, historical markets in promoting increased productivity, social change and growth. He also attempts to integrate Sraas
appreciation of the multi-sectoral nature of output into a dynamic model
of the economy.
Failures in growth also attract Post Keynesian attention, with the most
notable failure being the Great Depression. The most cogent explanation
Growth theory 179

here has come from Hyman Minskys blending of Irving Fishers debt
deflation hypothesis with Keyness analysis of the formation of expectations under uncertainty. Minsky developed the hypothesis that a period of
stable economic growth will lead to capitalists and bankers revising their
risk aversions, leading to an increased willingness to take on debt to finance
expansions that will inevitably result in a period of financial stress and a
collapse in the growth rate. This hypothesis that stability is destabilizing
neatly returns us to the Harrodian foundations of Post Keynesian growth
theory, by providing a link between the rate of growth and the development
of capitalist expectations and hence their rate of investment.
Nonlinearity and chaos theory (or more properly complexity theory)
are playing an increasing role in modern Post Keynesian work on growth.
Nonlinear relations arise naturally in economics out of interactions
between variables (for example, by the multiplication of the wage rate times
the level of employment to determine the wage bill) and obvious nonlinear
social relations (such as the relationship between the rate of growth of the
economy and profits and the willingness of capitalists to invest). When put
into mathematical models of growth using diererence or dierential
rather than simultaneous equations these nonlinearities in turn generate
the stylized fact that distinguishes the Post Keynesian approach to growth
theory from the neoclassical: unstable, cyclical growth. It is thus possible to
have models of the long run in which the system continues to fluctuate, and
in which the system never converges to an equilibrium. Chiarella and
Flaschel (2000) is a very sophisticated example of this approach, using
building blocks that would be acknowledged by traditional Keynesians as
well as Post Keynesians.
These models can be partially characterized by the mathematical properties of their equations, but modern computer technology has also added
the possibility of numerically simulating the behaviour of complex, highdimensional models with far-from-equilibrium dynamics. This technology
obviates the need to make simplifying assumptions, such as constancy of
income shares, which were previously used to make dynamic reasoning
tractable.
A comprehensive Post Keynesian theory of growth would clearly involve
the following elements: a treatment of the causes of innovation in a market
economy, where competition is primarily in product dierentiation rather
than price; multiple industry sectors rather than the abstraction of homogeneous output, so that the disproportional growth of the real economy is
mirrored by the model, and the impossibility of producing an aggregate
measure of capital is explicitly acknowledged; a relationship between the
rate of technical progress and the income distribution and eective demand
constraints inherent in a capitalist economy; a key role for non-neutral
180
Growth theory
monetary factors, with the possibility that debt accumulation dynamics

may on occasions retard and even reverse the process of growth; and a
resulting model that generates both endogenous cycles and endogenous
growth, with a significant possibility of economic breakdown under the
weight of financial factors. While the many strands of Post Keynesian
thought to date have provided most of the necessary strands, it remains true
that blending an overall tapestry remains a research project for future Post
Keynesians.
S K
See also:
Business Cycles; Dynamics; Financial Instability Hypothesis; Growth and Income
Distribution; Innovation; Kaldorian Economics.
References
Blatt, J. (1983), Dynamic Economic Systems: A Post Keynesian Approach, Armonk, NY: M.E.
Sharpe.
Chiarella, C. and P. Flaschel (2000), The Dynamics of Keynesian Monetary Growth, New York:
Harrod, R.F. (1939), An essay in dynamic theory, Economic Journal, 49 (193), 1433.
Kornai, J. (1990), Vision and Reality, Market and State: Contradictions and Dilemmas
Revisited, New York: Routledge.
Kregel, J.A. (1975), The Reconstruction of Political Economy: An Introduction to PostKeynesian Economics, London: Macmillan.
Nell, E.J. (1998), The General Theory of Transformational Growth, New York: Cambridge
University Press.
Income Distribution
The dominant theory of distribution in modern economics is the neoclassical marginal productivity approach, also known as the supply and
demand approach. Marginal productivity theory conceives capital as a
productive factor, and it argues for the existence of an inverse monotonic
relation between the rate of profit and the quantity of capital employed in
the production process. This relation constitutes the demand for capital
schedule. The supply of capital is determined by households portfolio
demands for capital, and the equilibrium rate of profit and quantity of
capital are then determined by the intersection of the supply and demand
curves for capital.
The process of wage determination is entirely analogous to that determining the rate of profit. Labour is also viewed as a productive factor, and
there exists an inverse monotonic relation between the wage rate and the
quantity of labour employed. This relation constitutes the demand for
labour schedule. The supply of labour is determined by households utilitymaximizing choice over leisure and market income, and the equilibrium
wage rate and employment level are determined by the intersection of the
supply and demand curves for labour.
Perfect competition is the hallmark of the marginal productivity theory
of income distribution. Departures from perfect competition can be introduced to explain such phenomena as discrimination in labour markets. This
introduces economic and monopoly rents, with some factors being paid
more than they would in a competitive market. This is also the neoclassical
approach to trade unions. However, these modifications retain the basic
marginalist approach to income distribution, interpreted as the outcome of
an exchange process based on choices at the margin in a world in which production is described by a continuous concave function that is homogeneous of degree one.
The concept of a production function is crucial to neoclassical theory,
providing the basis for marginal products from which are derived the
demand for labour and capital schedules. The logical foundations of this
concept formed the initial focus of a debate that was to become known as
the Cambridge capital controversies. The controversy was launched by
Joan Robinsons (195354) article challenging the existence of an aggregate
production function on the grounds that it is impossible to aggregate
heterogeneous capital. The Robinson critique has now been largely
accepted, but is generally ignored by neoclassical economists in practice.
181
182
Income distribution
Another criticism of neoclassical marginal productivity theory, which is

post-modernist in character, emphasizes the social construction of marginal products (Palley 1996a, pp. 647). Within existing accounts of marginal productivity theory, factors are paid their marginal products, which are
objectively measurable. Such a position assumes that objective measurement is possible. Yet measurement is intrinsically social, being an act of
interpretation based upon socially negotiated rules. These rules attribute
value and are derived from understandings that are themselves socially
derived. As knowledge, beliefs and social arrangements change, so too will
measurements. Who does the measuring aects the measurement outcome.
Such considerations introduce a radical subjectivism into neoclassical production theory that parallels ordinal utility theory, which introduced
radical subjectivism into neoclassical consumer theory in the 1930s. As a
result, even if well-defined production functions exist, income distribution
can never be the result of a purely technical process and is always inevitably tainted by social forces.
A key feature of the neoclassical supply and demand approach is that it
is a joint theory of employment and factor price determination. Supply and
demand schedules determine both prices and quantities, and the downward-sloping labour demand schedule imposes a binding trade-o whereby
real wages can only increase if employment falls. The neo-Ricardian framework, developed by Piero Sraa, aims to sunder the link between wages and
employment. The Sraan system has a number of appealing properties.
The determination of the normal wage reflects social and historical forces,
opening the way for the introduction of bargaining power concerns. It also
breaks with the labour demand curve notion that the level of real wages
constrains the level of employment. Instead, in the Sraan system the real
wage constrains the profit rate, and the binding trade-o is between the
profit rate and wages. Since the model does not use aggregate capital, but
instead only requires a competitively maintained common rate of profit on
the value of inputs, it is not subject to Robinsons (195354) capital critique.
The traditional Marxian approach to income distribution is constructed
through the lens of the labour theory of value, and the focus is on the extraction of surplus value. In this framework, concern lies with the rate of surplus
value, which measures the degree of exploitation of labour. Over the last 25
years, a distinctive American neoclassical school of Marxism has developed, leading figures of which are Samuel Bowles, Herbert Gintis and John
Roemer. These economists accept the existence of a well-defined neoclassical production function, but they break with the neoclassical assumption
that technology is exogenous. Instead, they argue that it is endogenously
selected. The significance of this argument is that choice of technology now
involves human agency, social context and control.
Income distribution
183
The importance of control for distributional outcomes signals the

importance of power. This brings into play the issue of perfect competition, which is another assumption embedded in neoclassical marginal productivity theory. Perfect competition ensures that both capital and labour
have no power. It is not that the two are equally powerful, but rather that
neither has any power. Removing the perfect competition assumptions of
costless mobility and perfect free information restores power to centre
stage.
The above neoclassical Marxian concerns link with the macroeconomics
of Michal- Kalecki (1942). A central component of Kaleckis macroeconomics is the mark-up, and its determination constitutes a key element of
the Kaleckian research programme. In the standard Kaleckian model
output is produced through a linear production function involving labour,
and prices are a mark-up over average cost. The mark-up determines the
wage and profit shares, bringing to the fore the question of what determines
the mark-up. A modern neoclassical industrial organization perspective
would focus on the degree of monopoly in product markets. Neoclassical
Marxism focuses on control and bargaining power issues.
Keynesian economics emphasizes the significance of aggregate demand,
and aggregate demand considerations figure centrally in the Post
Keynesian approach to income distribution. The Post Keynesian approach
was developed by Kaldor (1956), and has its roots in another side of
Kaleckis (1942) macroeconomics. Rather than the mark-up, the pivotal
point is dierent propensities to consume out of wage and profit income.
Given exogenous propensities to save out of profit and wage income, the
profit share is determined exclusively by the investment share of output. If
the propensity to save out of wage income is zero, which is Kaleckis
assumption, then the profit share depends on just the investment share and
the propensity to save out of profits. Higher investment spending raises the
profit share, while a higher propensity to save out of profits lowers it. This
leads to the Kaleckian dictum that capitalists earn what they spend, while
workers spend what they earn. The logic behind this Post Keynesian result
is that investment needs to be financed by saving, and income distribution
must therefore be appropriate to support the right level of saving. If investment spending goes up, a higher profit share is needed to generate additional saving. If the propensity to save goes up, a lower profit share is
needed to reduce total saving.
In Kaldors (1956) Post Keynesian model there is an implicit class structure consisting of workers and capitalists. Pasinetti (1962) explicitly models
this class structure, using the assumptions that (i) workers have a lower propensity to save than capitalists, (ii) capitalists only source of income is
profit income, (iii) workers receive both wage and profit income wage
184
Income distribution
income for supplying labour and profit income on their saving, and (iv) the
rate of interest is equal to the rate of profit. The assumption that workers
have a lower propensity to save ensures that their saving out of wage and
profit income does not drive down the capitalists ownership share of the
capital stock to zero. Given these conditions, Pasinetti shows that steadystate income distribution is unaected by workers saving behaviour. The
economic logic is simple. In a steady state the capitalist share of the capital
stock is constant, and they must save sucient to maintain this ownership
share. Consequently, the profit share must be such that it can support a
share of saving appropriate to maintaining the capitalists ownership share.
Viewed in this light, the Post Keynesian theory of income distribution
might better be thought of as a theory of wealth ownership.
Pasinettis theorem regarding the irrelevance of worker saving behaviour
has been remarkably robust with regard to introduction of other sources of
saving. It holds when government saving is introduced via the government
budget constraint and also in the presence of life-cycle saving behaviour.
However, there are a number of limitations to the Post Keynesian approach
to income distribution. First, it is an exclusively real theory of the interest
rate, which is determined by the profit rate, and this is at odds with
Keynesian theory, which emphasizes liquidity preference. Introducing
monetary factors into the analysis invalidates Pasinettis theorem regarding
the irrelevance of the workers propensity to save for steady-state distributional outcomes. To the extent that money balances are disproportionately
held by workers, they must save more to maintain their share of the money
stock, which influences steady-state income distribution. Palley (1996b)
introduces financial intermediation and inside debt, and demonstrates that
Pasinettis theorem holds if lending is done via a loanable funds market, but
is invalidated if done through a banking system with endogenous credit
money.
In sum, the Post Keynesian approach to distribution, with its focus on
the dynamics of capital accumulation within classes, makes a valuable contribution. Yet, despite introducing class, it makes no mention of class conflict in the form of labour market struggle. Nor is there any mention of
product demand conditions, in the form of the rate of capacity utilization.
These considerations suggest that is an incomplete account of the determination of income distribution.
The above reflections on the Post Keynesian approach point to the fact
that income distribution is likely determined by a complex of factors, suggesting the need for a synthetic approach. The neo-Marxian approach to
income distribution can be synthesized with Keynesian demand considerations. Palley (1998) presents a short-run neo-Marxian model in which
labour market bargaining conditions determine the distribution of income,
Income distribution
185
and aggregate demand conditions determine the state of labour market

conditions. Because aggregate demand is aected by income distribution,
owing to Kaleckian dierences in the propensity to consume out of wage
and profit income, there is a feedback loop between aggregate demand (the
goods market) and income distribution (the goods market). The canonical
long-run version of the neo-MarxianKeynesian model is attributable to
Goodwin (1967), who constructs a model in which labour market conditions drive profit rates, profit rates drive the rate of accumulation, and the
rate of accumulation feeds back to aect labour market conditions. When
placed in a multiplieraccelerator framework, this generates cyclical
growth, with a full-employment profit rate squeeze sending the economy
into a phase of slower growth with rising unemployment that lasts until the
profit rate has recovered.
The theory of distribution is more than just a matter of social and ethical
interest. It also profoundly aects the way in which we view the economy.
The neoclassical marginal product of labour is interpreted as the labour
demand curve, and this enforces an inexorable trade-o between wages and
employment. This trade-o drives opposition to minimum wages and
unions, and it also drives macroeconomic policy recommendations that
aim to lower unemployment by weakening employee protection and
making wages downwardly more flexible. Yet all of these policy stances are
predicated on a theory whose microeconomic foundations are deeply controversial. Moreover the validity of these policies is also questioned by
macroeconomic monetary analyses that show why lower real and nominal
wages may not increase employment.
The theory of distribution lies at the core of theories of output and
employment determination. Seen in this light, it provides a window on the
range of theories explaining the operation of modern capitalist economies.
That being so, it is startling that marginal productivity theory is the only
theory taught in most university classrooms. It is sometimes suggested that
neoclassical theory represents the culmination of grand theory in economics, and from here on in it is a matter of making small advances at the margin.
However, if economic theory is viewed as an exercise in story-telling, then
the stories economists tell will be influenced by the social and political environment. A change of environment could easily give rise to a burst of new
story-telling, and the grandest theory of all the theory of income distribution could then be subject to a wave of re-telling.
T I. P
See also:
Capital Theory; Growth and Income Distribution; Kaldorian Economics; Marginalism;
Sraan Economics; Wages and Labour Markets.
186
Ination
References
Goodwin, R.M. (1967), A growth cycle, in C.H. Feinstein (ed.), Socialism, Capitalism, and
Economic Growth, Cambridge: Cambridge University Press, pp. 548.
Kaldor, N. (1956), Alternative theories of distribution, Review of Economic Studies, 23 (2),
83100.
Kalecki, M. (1942), A theory of profits, Economic Journal, 52 (2067), 25867.
Palley, T.I. (1996a), Out of the closet: the political economy of neo-classical distribution theory, Review of Radical Political Economics, 28 (3), 5767.
Palley, T.I. (1996b), Inside debt, aggregate demand, and the Cambridge theory of distribution, Cambridge Journal of Economics, 20 (4), 46574.
Palley, T.I. (1998), Macroeconomics with conflict and income distribution, Review of
Political Economy, 10 (3), 32942.
Pasinetti, L. (1962), Rate of profit and income distribution in relation to the rate of economic
growth, Review of Economic Studies, 29 (4), 26779.
Robinson, J. (195354), The production function and the theory of capital, Review of
Economic Studies, 21 (2), 81106.
Inflation
The Post Keynesian theory of inflation is eclectic in a way that the neoclassical or orthodox theory is not. It allows for multiple causes of, and explanations for, inflationary phenomena, as opposed to the typically mono-causal
nature of orthodox theory. The main reason for this is a more realistic view
of the credit-creation process, namely the theory of endogenous money
(Moore 1979, Wray 2001). This implies that any factor tending to raise
money costs has the potential to cause an increase in the general price level,
as firms/entrepreneurs incurring the costs will have access to newly-created
financial resources to pay for them. Orthodox theory, with an exogenous
money supply, only allows for changes in relative prices as long as the money
supply remains fixed. If one element of costs increases, this must be oset by
a fall elsewhere.
The backbone of orthodox theory is one version or another of the quantity theory of money, illustrated by the equation of exchange, MVPY.
This is an identity in principle, but if it is assumed that the money supply (M)
is exogenous (controlled by the central bank), the velocity of circulation (V)
is roughly a constant, and that money is neutral and superneutral (so that
changes in money-supply growth will not aect the growth rate of real GDP
(Y)), it also provides a simple theory of the aggregate price level (P). Letting
lower-case letters represent proportional rates of change, we have:
pmy
(1)
The inflation rate (p) will be determined by the rate of growth of the money
supply (m) minus the rate of growth of GDP (y). Although the theory of
Ination
187
monetarism, associated with the work of Milton Friedman in the midtwentieth century, was more sophisticated than this (allowing for variable
velocity and short-run output eects), this statement captures the basic
idea. It is a mono-causal theory of inflation in the sense that almost all variation in inflation is attributed to variation in monetary growth. There are
some obvious problems, however, such as how precisely to define the money
supply in an era of rapid financial innovation, and whether or not it can
sensibly be regarded as exogenous in any reasonably sophisticated banking
system. One such problem, clear to Fed-watchers in financial markets, is
that in practice the monetary policy instrument is usually a short-term
interest rate (the Federal Funds rate in the United States), rather than any
quantitative measure of the money supply or monetary base.
Contemporary central bankers, therefore, seem to have a practical theory
of inflation owing more to Knut Wicksell than to Irving Fisher or Milton
Friedman, such as:
p(rn i)
(2)
where is a positive coecient. Here the nominal interest rate i has the connotation of the policy-determined rate set by the central bank, and rn is
Wickells natural rate, a real interest rate supposedly determined in the
market for real capital, independently of any monetary influence. If the
policy-determined rate is set too low this creates an incentive for bank borrowing as long as the discrepancy exists, hence an endogenous increase in
the rate of money growth, and ultimately inflation. The reverse occurs if the
rate is set too high. Wisdom in monetary policy entails searching for the
correct setting of the policy-determined rate, to precisely match the natural
rate. Then, supposedly, there would be no inflation, and the unemployment
rate, GDP growth rate, and the interest rate would all be at their natural
levels. The contemporary Taylor rule for monetary policy, for example, can
plausibly be interpreted in these terms. As with the original Wicksellian
model, this approach provides a twist on the quantity theory in the sense of
a more realistic conception of how money is introduced into the economy.
However, it does not depart too far from orthodoxy, due to the underlying
assumption that the economy always tends to a non-monetary market equilibrium. Inflation/deflation is caused by the gap between the natural and
policy-determined interest rates, which can be created either by deliberate
monetary policy, or by a change in the natural rate itself, not matched by the
monetary authority. If this is not precisely a mono-causal view of inflation,
there are still only a limited number of possible inflation sources.
Contrary to the above, according to Joan Robinson (1979, p. xix) one of
the most important insights of the Keynesian revolution was . . . that the
188
Ination
general level of prices in an industrial economy is determined by the general

level of costs, and . . . the main influence upon costs is to be found in the
relation between money-wage rates and output per unit of employment
(emphasis added). This can be illustrated by the formula, PkW/A, a rival
to the orthodox equation of exchange, associated with the work of Sidney
Weintraub (Davidson 1994). Here W is the average nominal wage rate, A
is average labour productivity, and k is the mark-up or profit share.
According to this, prices will rise if the money wage rises, if the mark-up
rises, or if productivity falls. If, further, the profit share stays roughly constant, this gives the theory of inflation alluded to by Robinson, that inflation is caused mainly by a rate of increase of money wages faster than
productivity growth. This can be written:
pw a
(3)
where w is wage inflation, and a is productivity growth. The remedy for

inflation, then, would not be changes in the rate of money supply growth
or the interest rate, as suggested by monetarists or Wicksellians, but specifically an incomes policy of some kind. In other words, regulations or agreements restricting the rate of growth of money incomes. Such policies, of
course, have their own problems of implementation, including the need to
gain public support for the restrictions. In particular, a concern of labour
unions where incomes policies have been suggested is that wage and price
controls should not turn out to be wage controls only. There should be a
measure of equity in controlling the receipts of other income groups also.
Some Post Keynesians have therefore responded with various clever proposals (Davidson 1994, p. 149) to meet these concerns, such as the taxbased incomes policy (TIP) suggested by Weintraub and Henry Wallich,
the market anti-inflation plan (MAP) of Abba Lerner and David Colander,
and, more recently, employer of last resort (ELR) proposals advocated by
members of the contemporary neo-Chartalist school, whereby government
employment at a fixed wage creates a buer stock programme for labour
(Wray 2001).
As mentioned, in the orthodox view of the world cost-based inflation
would be ruled out by the idea that the money supply is fixed, or at least
under the control of the central bank. In Post Keynesian theory the
assumption is that the money supply will normally increase endogenously
via credit creation to accommodate or validate any underlying increase in
costs. This point can be illustrated by combining the expressions MVPY
and PkW/A, noting that YAN. If both velocity and the mark-up are
held constant we obtain:
mwn
(4)
Ination
189
where wn is the growth rate of the nominal wage bill. So from this, what
seems to be mainly at issue is the question of causality. Reading equation
(4) from right to left the inference is that an increase in the growth of the
wage bill causes the money supply to grow (and hence causes inflation). The
monetarists might put this the other way around, arguing that an exogenous increase in money-supply growth first causes price inflation, and only
later causes wage inflation in a catching-up process. The debate comes
down to which is the more realistic description of banking practices in the
contemporary credit economy. One item on the Post Keynesian side of the
ledger is that, if money supply growth is just a link in the chain, as opposed
to a primary causal factor, it becomes much less urgent to have a precise
statistical measurement of this magnitude, which in any event seems to be
increasingly dicult in current conditions.
Note that the Post Keynesian approach lays much stress on the influence
of income shares on the overall price level. In the simple example looked at
above there are just two shares, the wage share and a generic profit share.
For other purposes, it may also be useful to distinguish, for example, the
rentier share from that of entrepreneurial capital, or to introduce other
types of income classification. Hence, an important development in the
Post Keynesian literature has been the notion of conflict inflation, put
forward by such authors as R.E. Rowthorn and A.K. Dutt (Lavoie 1992),
which generalizes the notion of conflict over income shares. Each of the
dierent income-earning groups tries to improve its real share by increasing nominal claims on output, and these claims are facilitated by the possibility of credit creation. Inflation results because this is the only way to
reconcile the competing real and nominal claims. Note that in an open
system, changes in the terms of trade will also be highly relevant if (for
example) labours target for its real wage includes a substantial proportion
of foreign goods. Again, some sort of consensus over income distribution
would be needed to reduce inflationary pressures.
If we were to briefly summarize the dierences between the orthodox and
Post Keynesian approaches to inflation sketched out above, this might
revolve around the old distinction between demandpull and costpush
inflation. At one time this was a staple of the textbooks, but it is now
neglected. From this perspective, orthodox theory basically has output
determined on the supply side, with money prices (and hence inflation)
determined by demand. The simplest version of Post Keynesian theory,
however, reverses this, and has prices determined on the supply side by costs
(compare the quotation from Robinson above), with output/employment
determined on the demand side, via the principle of eective demand.
There is obviously some truth to this characterization of the debate, and
the Phillips curve explanation of inflation, postulating a trade-o between
190
Ination
inflation and unemployment to be exploited by demand-side policy, has

always had an ambiguous status among Post Keynesians (Davidson 1994;
Palley 1996).
However, it would be going too far to say that the Post Keynesian
school neglects the concept of demandpull inflation. The key underlying
assumption of endogenous money allows for an eclectic view in this
respect. The Post Keynesian model does allow for what Keynes called
true inflation, that is, inflation caused by continued increases in demand
after some putative situation of full employment has been reached. Here
the analysis of the inflationary gap from the old-fashioned textbooks
would apply, with a caveat perhaps as to how frequently such situations
occur in practice, and also that full employment in this sense must be
interpreted as a genuine measure of capacity utilization, rather than the
market-determined natural rate (of unemployment) of the neoclassicals.
In addition, as in some versions of conflict inflation theory, it can be recognized that feedback does exist between demand pressures and the
market power of dierent groups competing for income share. Some of
this may well be operative before overall full employment is reached, due
to bottlenecks of various types. Hence, it should be possible for Post
Keynesian theory not only to stress the costpush aspect neglected by
orthodox theory, but also to reclaim the demandpull approach (Palley
1996, p. 166).
Inflation is a complex social process, and it seems unlikely that there is any
one explanation of the phenomenon valid for all times and all places. For
any theory which asserts, for example, that higher growth is always associated with higher inflation, it is always possible to point to empirical
instances of the opposite, either stagflation (low growth with high inflation)
or, more benignly, non-inflationary growth. The Post Keynesian approach
may therefore ultimately have an advantage over more orthodox explanations of inflation, precisely because of its open-ended and eclectic nature.
The key features are money-supply endogeneity and the rejection of natural
rate concepts (either of the interest rate or of unemployment). These allow
for coherent explanations of most of the empirical possibilities. At a
minimum, the potential for costpush inflation is recognized, which in itself
must provide for a richer description of real-world events than a view in
which such things are ruled out by assumption.
J S
See also:
Endogenous Money; Full Employment; Money; Rate of Interest; Tax-based Incomes Policy;
Wages and Labour Markets.
Innovation
191
References
Davidson, P. (1994), Post Keynesian Macroeconomic Theory: A Foundation for Successful
Economic Policies for the Twenty-First Century, Aldershot: Edward Elgar.
Elgar.
Moore, B.J. (1979), Monetary factors, in A.S. Eichner (ed.), A Guide to Post Keynesian
Economics, White Plains, NY: M.E. Sharpe, pp. 12038.
Palley T.I. (1996), Post Keynesian Economics: Debt, Distribution and the Macro Economy,
London: Macmillan.
Robinson, J. (1979), Foreword, in A.S. Eichner (ed.), A Guide to Post Keynesian Economics,
White Plains, NY: M.E. Sharpe, pp. xixxi.
Wray, L.R. (2001), Money and inflation, in R.P.F. Holt and S. Pressman (eds), A New Guide
to Post Keynesian Economics, London: Routledge, pp. 7991.
Innovation
As the research field of economics deepened over more than two centuries
since the Industrial Revolution, the linkage between innovation and economic development, which early classical writers emphasized, has become
more tenuous. Only economists examining the economy as a vast interconnected open systems canvas continued to maintain this link, notably Karl
Marx, Rosa Luxemburg, Michal- Kalecki and Joseph Schumpeter. In the
1990s this situation altered dramatically, with an enormous expansion of
research into innovation from many perspectives. Post Keynesian economics, after some early eorts in this area, has tended to neglect this issue.
Innovation can be defined as the application of knowledge in a new form
to increase the set of techniques and products commercially available in the
economy. These techniques can be technological or organizational. The
forms that innovation can take are (i) continuous incremental (or Kaizen);
(ii) radical discontinuous based on research and development (R&D); (iii)
technological systems change based on a cluster of innovations; and (iv)
techno-economic paradigm shift due to major structural change (for
example, the steam engine, information technology). Forms of innovation
can dovetail into higher-order innovation, thus becoming increasingly more
important to society.
Contemporary research into innovation has taken two approaches. One
is the study of broad-based evolutionary change in the long-term structure
of capitalism, while the other is narrow-based entrepreneurship studies at
the firm and industry levels. Post Keynesian analysis links innovation to
investment decision making, so that the elements of eective demand and
cyclical volatility at the broad base are related to the cumulative processes
in all forms of innovation at the firm/industry level. This entry focuses on
analysis of innovation that is based on this Post Keynesian perspective and
incorporates research from both approaches.
192
Innovation
Attempts to incorporate investment into the theoretical analysis of innovation have been limited. Two major exceptions to this are Salter (1960)
from the neoclassical perspective, and Freeman and Perez (1988) from the
evolutionary perspective. Both innovation studies set up economic snapshots which provide case study patterns to show the plausibility of the
theoretical relations they derive with respect to investment.
Salter examines technical change and its implications for means of production (MOP) increments at the margin in dierent industry sectors. In an
exceptionally insightful manner, Salter recognizes the gap between available innovation and its application via investment. He uses market signals
to indicate possible postponements in the use or introduction of more innovative MOP and consequent delays in scrapping old MOP; thus the capital
stock becomes fossilised (Salter 1960, p. 154). This exposes technical
change to dierent rates of productivity growth between industries. The
leading Post Keynesian, Georey Harcourt, used Salters approach to technical change in a number of significant articles in the 196575 decade, culminating in Harcourt and Kenyon (1976) on the impact of investment
decisions incorporating technical change on pricing behaviour.
Freeman and Perez (1988) take a dynamic structural adjustment view of
the economy with respect to innovation, and note the mismatch of current
investment to new available technology. Rather than market signals, they
emphasize variations in the climate of confidence related to the type of
innovation and the life cycle of the industries which account for this mismatch, leading to intensified investment instability. Courvisanos (1996) has
extended this work in an eort to incorporate life-cycle innovation into
investment instability.
The classic proposition of an investment model with innovation comes
from Joseph Schumpeter, who recognized that the investment function
responds to waves of optimism and pessimism that create clusters of innovation, and thus bunching of investment. This produces susceptibility to
unstable investment cycles and the development of a trigger mechanism to
initiate fundamentally new innovation systems with long-wave implications. Kalecki endorses and reinforces this cycle-trend eect that innovation has on the investment function. The intensity of innovation both
aects the amplitude of investment cycles and shifts the trend path of
investment growth, by flows of vicious and virtuous circles. Virtuous circle
eects occur as innovation intensity rises, increasing the amplitude of the
upper turning-point of the investment cycle and shifting the trend path
upwards. Vicious circle eects increase the amplitude of the lower turningpoint, shifting the trend downwards. The pace of innovation is a shift
parameter in the Kaleckian investment function.
The cause of clustering of innovation and subsequent bunching of
Innovation
193
investment (clust-bun) is in debate. The Kaleckian feature of expanded

reproduction has not been recognized by the protagonists in this debate.
The prerequisite for clustering is deep depressions or breakthroughs in
technology, both reflecting reactions by the private and public sectors to
deep problems in the downswing of the previous business cycle. Then, the
bunching requires eective demand stimulus through widespread diusion
of the cluster eect that can only be achieved through the availability of a
surplus for investment (private profits and public deficit spending).
Impediments to this clust-bun eect reside in the institutional frameworks
of nations, particularly those with still dominant mature industries with
older technologies (Freeman and Perez 1988, pp. 5865). Increased uncertainty arising from large investment in the new technology systems also
adds a further impediment through increased macroeconomic volatility,
slowing down the diusion process.
The causal sequencing of innovation and investment is reversed in work
done by Nicholas Kaldor and Joseph Schmookler, with the rate of investment determining the rate of innovation. Kalecki also recognizes this
sequence, despite having identified the innovation-driven process. Kalecki
places this investment-driven process clearly into an appropriate context by
viewing the innovation process as part and parcel of ordinary investment (Kalecki 1954, p. 158), or endogenous innovation.
Instead of unidirectional causality, the discussion above clearly implies
a circular flow, where one innovation process feeds into the other. Kaldors
principle of cumulative causation is the self-reinforcing dynamic in the
circular process of investment demand leading to innovation that then
stimulates further investment. The distinction between exogenous and
endogenous innovation specifies how innovation enters this cumulative
causation process. In this context, R&D expenditure is central to the endogenous innovation process, with large firms with strong profit results having
the ability to undertake large R&D spending, while registration of patents
from R&D eorts reflects the clustering of innovations.
Gomulka et al. (1990, p. 535) attempt to provide ergodic closure to the
Kalecki trend and cycle theory. They argue that Kaleckis central role of
innovations in preventing the trend rate of unemployment from increasing
is unsupportable, as the balanced growth rate which Kalecki took to be
stable is, in fact, unstable, rendering it unsuitable to serve as the trend
growth rate. Lavoie (1992, pp. 297327) examines Kaleckis innovation and
investment analysis at the theoretical level and rejects the ergodic closure
assumption which ties this theory back to the neoclassical mainstream.
Kalecki clearly assumes that the rate of capacity utilization may diverge
from its full-capacity rate even in the long run, with the reserve army of the
unemployed as a typical feature of capitalism for a considerable part of the
194
Innovation
cycle. This asserts instability, as the dynamic non-ergodic business cycle has
innovation-creating conditions that move the trend growth away from any
analytical stability.
In Kaleckis view of innovation, endogenous innovation is of secondary
importance from the scientific standpoint, coming as it does from: (i) slight
adaptations of previous capital equipment; (ii) cosmetic improvement in
old products; and (iii) extension of previous raw material sources. Such
innovation is called endogenous because it is the cycle itself that induces the
innovation and, with it, higher levels of investment orders. With endogenous innovation occurring in a Kaleckian macro economy, the analysis can
focus on how such innovation is developed at the firm/industry level and
consequently aects the economy.
The firms R&D expenditure is a form of intangible investment to be
incorporated in the long-term business investment plan. This enables the
firm to hold a stock of innovations that are ready to be applied when susceptibility to investment risk is relatively low. In this way endogenous innovation can be generated and directed by a process of investment. When a
firm decides to increase investment at relatively low susceptibility under
competitive pressures and higher costs of postponement, its R&D investment in the past makes these innovations ready to implement. R&D investment eectively increases the strategic productive capacity of the firm. In
an industry where innovation is a regular competitive strategy, R&D expenditure would be large and would vary under the same susceptibility pressures as capital expenditure. In an industry where innovation is only
occasionally implemented, R&D expenditure would be small and relatively
constant over the investment cycle.
The endogenous creation of innovations out of low susceptibility makes
some MOP obsolete and thus not part of excess capacity calculation. Also,
oligopolistic firms (and industries) lobby for the assistance of governments
in reducing private costs of production (through subsidies, tax concessions
or protection) when these firms attempt to expand their market by innovations in order to utilize new, and decommission old, idle productive capacity. Such innovation and underwriting of the related risks reduces the rate
of increase in susceptibility and encourages an investment recovery.
R&D amounts in aggregate to a large body of investigation going on
continuously (at dierent rates of intensity). This large R&D spending and
related innovation eects are bound to lead to some major new discovery
or invention which is related to the total aggregate R&D, rather than to
any particular R&D project. This discovery is linked to possible small
developments in various laboratories and informal networks between firms
and industries, eventually coming to fruition in some way divorced from
any specific competitive behaviour. New technological paradigms come out
Innovation
195
of such aggregate developments and are the basis of structural change to a

new long wave of boom and prosperity (Freeman and Perez 1988,
pp. 4758). Changes in technological systems and paradigms arise only
after all the minor improvements (endogenous innovation) are squeezed
out of the old systems and paradigms by monopoly capital entrepreneurs
who want to protect existing MOP and delay the new paradigm taking over.
There is also a log jam in endogenous innovations based on the new paradigm, which compounds the latters slow initial adoption. This occurs
when established powerful entrepreneurs, with much old MOP, cannot
justify the entire shake-up of industries, since not enough interrelated clusters have been formed.
Technological paradigm shift leads to exogenous innovation input aecting the investment cycle. Introduction of a new paradigm produces a large
exogenous boost to industry investment at low susceptibility points. This
investment boom relates to paradigm changes in large important industry
sectors that adopt new technology systems (for example, petrochemical
innovations), or in the whole economy (for example, steam engine innovations). Either way, the investment boom is strong and resilient over a series
of future cycles in susceptibility.
As the institutional framework slowly adapts to the new technological
system, entrepreneurs reactions to uncertainty of profits result from competitive pressures and growing ineciencies of old MOP. This induces
adaptation (by industries) and imitation (within industries) of technological trajectories that are totally new, establishing, at very low susceptibility,
the new investment upturn. This creates a new investment boom and at the
same time re-establishes the conditions for a new phase of steady development. A paradigm shift occurs when the newly-adapted technological
systems pervade the whole economy. Many from the evolutionary school
identify such a shift with the beginning of a new long wave in the economys
development.
This analysis links together the two types of innovations described by
Paul Baran and Paul Sweezy, namely normal (or endogenous) and epochmaking (or exogenous). A period of secular decline in economic development can now be associated with the limitations of scale production in
oligopolistic competition, as the old technology systems are running out of
possible new adaptations. Diusion of the old systems through endogenous innovation slows down, and imitators become considerably fewer. The
large powerful corporations attempt to protect existing capital values and
ignore the new technological systems that are being developed on the
fringes of the corporate world. This tends to exacerbate the mismatch
between new technologies and a powerful institutional framework based
around monopoly capital. It was Steindl, back in 1952, who recognized this
196
Institutionalism
secular decline as the incentive to reduce surplus capacity and invest in

established monopoly capital sectors. In his 1976 introduction to the 1952
book reprint, Steindl stated that he was ready to admit a possibility which
I denied in my book: that it might be the result of exhaustion of a long technological wave (1976 p. xv). In this way, the conclusions of the Kaleckian
and evolutionary traditions can be integrated.
J C
See also:
Business Cycles; Growth Theory; Institutionalism; Investment; Kaldorian Economics;
Kaleckian Economics.
References
Courvisanos, J. (1996), Investment Cycles in Capitalist Economies: A Kaleckian Behavioural
Contribution, Cheltenham, UK and Brookfield, VT, USA: Edward Elgar.
Freeman, C. and Perez, C. (1988), Structural crises of adjustment, business cycles and investment behaviour, in G. Dosi, C. Freeman, R. Nelson, G. Silverberg and L. Soete (eds),
Technical Change and Economic Theory, London: Pinter, pp. 3866.
Gomulka, S., A. Ostaszewski and R.O. Davies (1990), The innovation rate and Kaleckis
theory of trend, unemployment and the business cycle, Economica, 57 (228), 52540.
Harcourt, G.C. and P. Kenyon (1976), Pricing and the investment decision, Kyklos, 29 (3),
44977.
Kalecki, M. (1954), Theory of Economic Dynamics, London: George Allen & Unwin.
Lavoie, M. (1992), Foundations of Post-Keynesian Economic Analysis, Aldershot: Edward Elgar.
Salter, W.E.G. (1960), Productivity and Technical Change, Cambridge: Cambridge University
Press.
Steindl, J. (1976), Maturity and Stagnation in American Capitalism, 2nd edition, New York:
Monthly Review Press.
Institutionalism
Institutionalism is an approach to economics that sees economic life as
taking place within a social context. In contrast to neoclassical economists,
institutionalists see human behaviour as determined more by social factors
than by deliberative individual thought.
Behaviour depends on the habits, the routines, and the customs of economic actors. These actors are households, workers and business firms, as
well as the government and its policies and regulations. Their habits, routines and customs are the rules they use to make decisions. They are matters
of law or tradition; and they get passed along by example, by societys
expectations, and by the power of the state. People tend to follow these rules
because they see everyone else doing so. People also follow these rules
because they provide a simple way to deal with the uncertainty and the
complexity of everyday life (Hodgson 1988).
When most people follow institutional rules, behaviour becomes more
Institutionalism
197
certain and individuals are less likely to feel foolish by acting dierently or
by being wrong when they make important choices. Institutional rules thus
provide for stability in a world of uncertainty.
Post Keynesians believe that methodological individualism prevents
economists from seeing the impact of social phenomena on individual
choice and also keeps them from providing useful guidance to society. By
adopting an institutionalist or social perspective on individual choice, Post
Keynesians can analyse how the perceptions of economic agents are
moulded by institutions and habits. It also lets them address important economic and social issues that are assumed away by neoclassical theory.
Institutions appear at several key places in Post Keynesian economic
analysis. First, institutional factors help explain the consumption behaviour
of households. Second, institutions help us understand the investment decisions of business firms. Third, institutional considerations lead to the creation of money in capitalist economies and are responsible for the unique role
of money. Finally, they help us understand the stability of capitalism, and
how and why economic policy can improve economic outcomes. We consider these items in turn.
Institutionalists see consumer preferences and consumer spending stemming from learned social behaviour rather than from any innate utility
functions. Consumer spending is determined by what is necessary to maintain a lifestyle similar to ones friends and neighbours, and possibly a lifestyle that is a bit more lavish than that of ones friends and neighbours. This
argument goes back to the work of Veblen (1899 [1908]), regarded as one
of the founding fathers of institutionalist thought.
These behavioural dispositions help explain why consumption is stable
and also why fiscal policy is able to expand or contract the economy.
Consumption is stable because it depends on spending habits. For most
middle-class households, this means spending most, if not all, of ones
regular pay-packet. As a result, the propensity to consume additional
income will be stable over time and also have a relatively high value.
A stable and high propensity to spend means that consumers will usually
spend a large fraction of any extra money that they receive. This has important policy implications. Tax cuts, even temporary tax cuts enacted during
a recession, will increase individual spending nearly dollar for dollar; likewise government spending will increase income (and therefore spending) by
some large fraction of any additional state expenditures. For this reason,
fiscal policy can be counted on to aect the overall macro economy in a
fairly predictable manner.
On the neoclassical approach, business investment is a rational and
maximizing decision. Firms compare the costs of investing (interest lost
due to borrowing or employing retained earnings) with the benefits (future
198
Institutionalism
earnings discounted for the time value of money). If benefits exceed costs,
the firm will undertake the new investment project; otherwise there will be
no new investment.
For Keynes, business investment was not undertaken on this basis,
because firms are unable to make the calculations required of them. Future
earnings are inherently uncertain. Post Keynesian economists also stress
the uncertainty of the investment decision; such decisions must be made on
the basis of educated guesses, gut feelings, or animal spirits. These decisions arise not out of individual contemplation, but rather from a collective process where everyone watches what everyone else is doing.
Firms will invest if and only if a lot of other firms are investing and
greater investment seems to be a safe choice. In contrast, when animal
spirits are pessimistic, few firms will invest. Under these circumstances, any
single firm that invests will experience sluggish sales and low or non-existent
profits. Here, the investment decision will turn out to have been a mistake.
For this reason, negative expectations by some firms quickly translate into
negative expectations by most firms, and little investment takes place. With
little investment there will be high unemployment, expectations will remain
poor, and the economy will remain mired in recession.
Money, for Post Keynesians, is an institutional construct that helps
reduce uncertainty. Holding money reduces uncertainty for the firm
because workers must be paid with money and debts must be repaid with
money; it reduces uncertainty for people because households know that
they will be able to pay for necessities in the future with the hoarded money.
In addition, money (unlike stocks and other assets) does not change much
in value from day to day and from month to month. By holding money,
households will not be subject to sharp declines in net worth in the future.
Because it is a refuge from uncertainty, people and business firms will
want to hold or hoard money in dicult economic times rather than spend
it. But this demand for money creates macroeconomic problems that lead
to even greater uncertainty and greater demand for money.
According to Davidson (1994, p. 18), money has two essential characteristics that lead to unemployment in a world with an unpredictable and
unknowable future. Money helps create unemployment because it has zero
elasticity of production and because there are no substitutes for money.
The former characteristic refers to the fact that no one is hired to produce
money when people want money rather than goods. The second property
refers to the fact that there is no substitute for money to pay o debts; even
if the return to holding money falls to zero, people still need money. When
people fear for the future, they desire to hold money. But because no one is
hired to produce money, workers get laid o, businesses cannot sell goods,
and everyone is more fearful about the future.
Institutionalism
199
In Post Keynesian analysis, the state serves as an important institution

that can counteract other forces undermining spending and leading to
unemployment. It does this by employing fiscal and monetary policies to
help control unemployment. The state can also help by creating institutional structures that tend to stabilize the economy property rights, a
central bank which operates as a lender of the last resort, and stable international economic relationships.
In addition, and in contrast to more traditional views of the state,
Keynes argued that the state was itself an important economic institution.
As Skidelsky (1989) has argued, Keynes saw the state as a set of institutions
that would provide for public goods and benefits, the prime benefits of
which would be full employment and important public goods and services
that business firms were unwilling or unable to provide. It fills in for other
social institutions when these institutions fail.
There are several mechanisms by which the state can fulfil these institutional functions. First, the state helps convert uncertainty and discontinuity into calculable risk. It gives economic actors confidence that the future
will be like the past. The state provides the laws and regulations that are
necessary for capitalist production to take place. It also provides for stability and security in life. This includes monetary stability, exchange rate
stability, welfare benefits, old-age pensions and deposit insurance. For
example, deposit insurance, in conjunction with central banks operating as
a lender of last resort, reduces the likelihood of bank runs and financial
collapse. State welfare systems are institutions which recognize that the
market, the family and social networks sometimes are not enough to generate individual security. People may not spend if fearful of the personal
consequences of becoming unemployed. A viable social safety net alleviates this concern (Larson 2002).
Second, the state also provides an anchor for decision making. Firms can
have more confidence in their own investment decisions when they know
that the government will help maintain aggregate investment and full
employment. This greater confidence will, in turn, generate more private
business investment.
For Keynes, it did not matter how the state spent its money; what mattered was that the money got spent. In a much-quoted passage, Keynes
writes about the need for more houses, hospitals, schools and roads. But,
he notes, many people are likely to object to such wasteful government
expenditures. Another approach was therefore necessary. If the Treasury
were to fill old bottles with banknotes, bury them at suitable depths in
disused coal-mines which are then filled up to the surface with town rubbish
. . . private enterprise [would] dig the notes up again [and] . . . there need be
no more unemployment (Keynes 1936 [1964], p. 129).
200
International economics
Keynes (ibid., p. 378) preferred, however, a somewhat comprehensive

socialization of investment. What he was advancing here was government
spending policies to stabilize the aggregate level of investment in the
national economy. For Keynes, the state needed to run deficits and invest
in education, infrastructure, health care and so on during times of high
unemployment. And during boom times, the government would need to
reduce its investment spending and run budget surpluses. By following
these budgetary rules, the economy would be more stable, businesses and
consumers would face less uncertainty, and both groups would spend more
(Pressman 1987, 1995).
In brief, employed correctly, monetary policy and fiscal policy function
as uncertainty-reducing institutions. They give business firms the confidence to invest, knowing the chances are good that production from any
new plants will be sold at a profit. They also give consumers confidence in
the future, and keep them from hoarding money in fear of bad economic
times. Other institutional arrangements created by the state that tend to stabilize the economy will have similar beneficial eects.
S P
See also:
Agency; Consumer Theory; Economic Policy; Investment; Money; Uncertainty.
References
Davidson, P. (1994), Post Keynesian Macroeconomic Theory, Aldershot: Edward Elgar.
Hodgson, G. (1988), Economics and Institutions: A Manifesto for a Modern Institutional
Economics, Philadelphia: University of Pennsylvania Press.
York: Harcourt Brace.
Larson, S. (2002), Uncertainty, Macroeconomic Stability and the Welfare State, Aldershot:
Ashgate.
Pressman, S. (1987), The policy relevance of The General Theory, Journal of Economic
Studies, 14 (4), 1323.
Pressman, S. (1995), Deficits, full employment and the use of fiscal policy, Review of Political
Economy, 7 (2), 21226.
Skidelsky, R. (1989), Keynes and the state, in D. Helm (ed.), The Economic Borders of the
State, Oxford: Oxford University Press, pp. 14452.
Veblen, T. (1899 [1908]), Theory of the Leisure Class, New York: Macmillan.
International Economics
The conventional approach to international economics divides the subject
into two separate branches. The micro part, called international trade,
analyses the determinants of countries exports and imports, and the
eects of alternative trade policies on economic welfare (including income
201
distribution), using purely real or barter models. The macro part, called
international finance, analyses balance of payments adjustment and
exchange rate determination using aggregative models that emphasize
monetary and financial factors. The pure trade theory assumes that automatic financial and macroeconomic adjustment mechanisms eectively
ensure the conditions (balanced trade with full employment) under which
trade follows comparative advantage and all nations gain from free trade.
Post Keynesians reject this bifurcated approach to international economics, and especially the implied neutrality of monetary and financial
factors with regard to real trade. Although Post Keynesian analyses may
focus on either the trade or financial side of the subject, the Post Keynesian
approach emphasizes how international trade and financial relations
impact on each other (see Deprez and Harvey 1999). Especially, in line with
the general view of a monetary production economy in which the financing of economic activity has non-neutral, real eects, Post Keynesians deny
the existence of automatic adjustment mechanisms that maintain balanced
trade and full employment as assumed in the standard pure trade theory.
This opens up the door to theories that emphasize the causes and consequences of trade imbalances, and the real adjustments in income and
employment required to oset them.
These theoretical distinctions are of vital importance because of their
implications for trade and financial policies. The conventional argument
for mutual benefits to all countries from free trade, based on the theory of
comparative advantage, is rooted in pure trade models that assume balanced trade and full employment as well as capital immobility. If any of
these assumptions are dropped, the theory of comparative advantage
breaks down, and it can no longer be presumed that free trade policies are
always in a nations best interest (although positive Post Keynesian analyses of trade policies are poorly developed to date). On the financial side, the
absence of automatic monetary adjustment mechanisms implies the need
for activist government policies and international cooperative arrangements (such as managed exchange rates and/or capital flow restrictions) in
order to foster more balanced and mutually beneficial trade and to promote
global full employment.
One core Post Keynesian idea that links international trade and finance
is Joan Robinsons theory of international conflict over limited global
markets, which she called the new mercantilism (see Robinson 1978,
pp. 190222). In a world with inadequate aggregate demand and involuntary unemployment, countries often seek to run trade surpluses in order to
boost their own output and employment. Since not all countries can run
surpluses at the same time, the countries that succeed in obtaining them
eectively compel other countries to run deficits, which saddle the latter
202
countries with lower national incomes and higher unemployment rates

than they would otherwise have. Thus, export-led expansion in some countries comes at the expense of import-imposed contraction in others, or in
Robinsons colourful adaptation of Adam Smiths famous remark
export-led growth is a beggar-my-neighbour policy. This analysis of conflictive trade relations stands in marked contrast to the conventional view
of largely harmonious trade relationships a view which ignores the existence of demand-side limits to global exports (and which allows for conflict
only over the barter terms of trade).
The Post Keynesian approach necessarily includes critiques of conventional theories of automatic balance of payments adjustment. For example,
in the specie-flow mechanism of David Hume and David Ricardo, which
applies to a fixed exchange rate system, a trade surplus (deficit) leads to an
inflow (outflow) of monetary reserves (gold or hard currencies), which in
turn raises (lowers) the money supply and causes a rise (fall) in the price level
that makes a countrys products less (more) competitive, and hence reverses
the trade imbalance. Post Keynesians criticize this theory because (among
other things) they deny that the supply of money determines the price level
in a modern industrial economy, as well as because this theory ignores the
role of capital flows in financing trade imbalances. As long as countries with
trade surpluses run osetting capital account deficits (that is, become net
lenders) and countries with trade deficits run osetting capital account surpluses (that is, become net borrowers), overall balance of payments equilibrium can be sustained without eliminating trade imbalances.
Another type of automatic stabilization mechanism involves flexible
exchange rates. Traditional analyses developed before the 1973 collapse of
Bretton Woods presumed that countries with trade surpluses would have
appreciating currencies and countries with deficits would have depreciating
currencies, leading to the restoration of balanced trade. However, Post
Keynesians argue that flexible exchange rates are monetary variables that
are driven primarily by financial capital flows and asset market speculation,
and hence need not move in the right direction for balancing trade and,
even when they do adjust, exchange rate changes may not generate the
desired improvements in the trade balance due to low price elasticities or
osetting price changes. While these points are recognized by some mainstream economists, the implication that imbalanced trade will not follow
comparative advantages is emphasized only by Post Keynesians.
Post Keynesians who follow in the Kaleckian tradition emphasize the
feedback eects of international competition on domestic profit mark-up
rates, and hence on the distribution of income between profits and wages
(see the chapter by Blecker in Deprez and Harvey 1999). When a currency
appreciates (or domestic costs rise relative to foreign), oligopolistic firms
203
squeeze pricecost margins in order to price-to-market, which in turn

leads to a fall in the profit share with possible negative repercussions for
investment and growth (although this may be oset by a boost to domestic
consumption arising from higher real wages and labour income). When a
currency depreciates (or domestic costs fall relative to foreign), the opposite happens as domestic oligopolies are enabled to raise their pricecost
margins without losing market share, income is redistributed from wages to
profits, and the potential repercussions for investment and growth as well
as for consumption are all reversed. Outcomes in which a redistribution of
income towards profits is contractionary are known as stagnationist,
while outcomes in which such a redistribution is expansionary are known
as exhilarationist. Mainstream economists have recognized the flexibility
of profit margins in response to exchange rate fluctuations what they call
partial pass-through but they have not analysed the eects on income
distribution, aggregate demand and economic growth.
At the microeconomic level, Post Keynesians argue that trade generally
follows absolute rather than comparative advantages (see Milberg 1994).
There are two dierent versions of this approach. For trade in standardized
products, which can be manufactured in similar processes with comparable
quality in a large number of countries, exports are based on competitive
advantages in unit costs of production, principally unit labour costs (that
is, wages adjusted for productivity, or wages in eciency units). Thus, the
countries with the lowest unit costs in a certain product, taking into
account current wages and other direct input costs (for example, raw materials and energy), relative to the productivity of labour and other inputs,
and adjusted for prevailing exchange rates, will export that product, regardless of whether they have a true comparative advantage in it. In labourintensive industries where technology is standardized and productivity is
fairly uniform, only low-wage countries will export the products, especially
when capital is mobile and firms can locate production wherever production costs are lowest (see Brewer 1985). Thus, there is some truth to the
popular notion of low-wage competition, but only if it is understood in the
proper context (that is, wages are adjusted for productivity, products are
standardized and capital is internationally mobile).
However, there are many internationally traded goods for which neither
production processes nor product qualities are standardized. For these
goods, a few technological leaders have either absolutely superior (lower
cost) technologies, or else produce absolutely higher qualities of the goods,
than any other countries. In these industries, which include important
sectors such as aerospace, industrial machinery, computer software and
medical equipment, trade is determined by technological gaps that is, the
countries with the superior technology or product are the exporters, and all
204
other countries are importers (see Dosi et al. 1990). Relative cost factors
(such as wages or exchange rates) are not important in these sectors and
product lines. Of course, individual products can shift over time from being
innovative products traded according to technological gaps to standardized
products traded according to competitive advantages, in line with Raymond
Vernons (1966) product cycle theory. As a result, the small club of innovating countries (led by the United States in the postwar period, and joined
more recently by Japan and others) needs to keep inventing newer innovative goods (for example, supercomputers or biotechnology) in order to stay
ahead of the competitive curve (since such countries typically have high
wages, and therefore cannot compete in standardized manufactures). The
technology gap theory of trade in innovative products thus complements
the absolute competitive advantages theory for standardized goods, allowing for a fairly complete explanation of most international trade especially
in manufactures.
Other Post Keynesian views can be covered more briefly since they are
discussed elsewhere in this volume. To deal with the volatility of flexible
exchange rates as well as the deflationary biases in traditional adjustment
mechanisms for deficit countries, some Post Keynesians have advocated a
return to a Bretton Woods-like system of adjustable pegs, but accompanied
by a mechanism to shift the burden of adjustment to the surplus countries.
Paul Davidson (199293) calls for the establishment of an international
monetary clearing house, which not only would create an international
reserve asset (international monetary clearing unit, or IMCU), but also
would require surplus countries to spend their surpluses and thus impart
an expansionary bias to the global adjustment process. In addition, many
Post Keynesians have advocated policies to discourage destabilizing flows
of short-term capital and to prevent speculative attacks on currencies.
However, Post Keynesian views on such policies vary, with some advocating Tobin taxes on foreign exchange transactions while others call for more
direct forms of capital controls or financial regulations.
Finally, the theory of balance-of-payments-constrained growth focuses
on the long-term consequences if trade imbalances cannot be sustained
indefinitely and countries are eventually forced to balance their trade (or at
least, to restrict trade imbalances to levels that can be financed through sustainable net capital flows). This view, which is elaborated by McCombie and
Thirlwall (1994), assumes that the long-run adjustment to balanced trade is
eectuated mainly through changes in output quantities (income levels or
growth rates), not by changes in relative prices (real exchange rates). The
implication is that countries with slow export growth and high income elasticities of import demand are condemned to grow more slowly than their
trading partners if they are forced to balance their trade in the long run.
Investment
205
Note that, while the basic version of this model does assume balanced trade
in the long run, it does not assume full employment or that the adjustment
mechanisms that restore balanced trade are neutral or painless, and the
model can be adapted to allow for capital flows.
R A. B
See also:
Balance-of-payments-constrained Economic Growth; Bretton Woods; Exchange Rates;
Globalization; Tobin Tax.
References
Brewer, A. (1985), Trade with fixed real wages and mobile capital, Journal of International
Economics, 18 (1/2), 17786.
Davidson, P. (199293), Reforming the worlds money, Journal of Post Keynesian Economics,
15 (2), 15379.
Deprez, J. and J.T. Harvey (eds) (1999), Foundations of International Economics: Post
Keynesian Perspectives, London and New York: Routledge.
Dosi, G., K. Pavitt and L. Soete (1990), The Economics of Technical Change and International
Trade, New York: New York University Press.
McCombie, J.S.L. and A.P. Thirlwall (1994), Economic Growth and the Balance-of-Payments
Constraint, New York: St. Martins Press.
Milberg, W. (1994), Is absolute advantage pass? Towards a Post Keynesian/Marxian theory
of international trade, in M. Glick (ed.), Competition, Technology and Money: Classical
and Post-Keynesian Perspectives, Aldershot: Edward Elgar, pp. 220236.
Robinson, J. (1978), Contributions to Modern Economics, New York: Academic Press.
Vernon, R. (1966), International trade and international investment in the product cycle,
Quarterly Journal of Economics, 80 (2), 190207.
Investment
The Post Keynesian theory of investment begins with the work of John
Maynard Keynes and Michal- Kalecki in the 1930s. Keyness ideas about the
determinants of investment in A Treatise on Money (1930 [1971]) and The
General Theory of Employment, Interest, and Money (1936 [1964]) depart
from the neoclassical theory mainly by emphasizing expectations of the
profitability of investment spending and the expectations involved in the
determination of financial market prices. The fundamental formulation of
investment in neoclassical theory is that it is determined by the intersection
of a downward-sloping schedule of the marginal productivity of increasing quantities of capital equipment relative to a given amount of the other
factors of production, with an upward-sloping schedule of the communitys willingness to abstain from consumption to supply quantities of
capital at dierent rates of return.
In the Treatise Keynes argued that the value of new investment goods
would rise and fall relative to the cost of production of new investment,
206
Investment
spurring changes in the level of investment, as the public and the banking
system changed their opinions about the desirability of moving wealth
between deposits and securities. A change in the willingness of the community to supply capital, that is, to save, would thus only change the level of
physical investment if it also changed the level of financial investment in
securities. The level of investment could change in turn, relative to the willingness to save, as the desirability of holding securities changed.
In the General Theory Keynes described the investment demand schedule as a schedule of the marginal eciency of capital. Although Keynes
held that his marginal eciency of capital schedule was equivalent to Irving
Fishers derivation of the rate of return over cost from neoclassical optimizing behaviour, he was clear that his schedule stood for the expected
profitability of additions to the capital stock and explicitly rejected the idea
that the value of capital was determined by the productivity of capital. This
schedule slopes downwards, Keynes wrote, since increased demand for
capital goods raises the cost of producing them and an increased supply of
any type of capital reduces its prospective yield. Later theorists were to say
that as these factors influencing the profitability of additional capital were
aected by the time period in which the capital is to be produced and
installed, the schedule should instead be called the marginal eciency of
investment, or additions to the capital stock per unit of time.
Keynes saw the supply of finance for investment as coming from the willingness of the public and the banks to give up liquidity, which determined
the relevant interest rate. The degree of excess bearishness, which determined the value of investment in the Treatise, became divided into the
expectations of profitability of the marginal eciency of capital schedule
and the degree of liquidity preference for holding money versus long-term
debts, given the quantity of money supplied by the central bank.
Kalecki (1990, 1991) criticized Keyness theory of investment on the
grounds that it was insuciently dynamic. That is, Kalecki questioned the
idea of having investment determined by the intersection of a given marginal eciency of capital schedule and the relevant mix of interest rates,
because he held that changes in the level of investment so determined would
feed back upon the marginal eciency of capital schedule itself, first as, for
example, increased investment increased aggregate demand and so the profitability of investment, and later as the new capital produced by the investment became available and so depressed the profitability of further
investment.
Kaleckis depiction of how investment spending is determined thus
requires a dynamic process in which investment interacts with output,
profits and the level of the capital stock, as in the flexible accelerator or
capital-stock adjustment model, though his own models were specified in
Investment
207
terms of investment, capital and profits, rather than output. Kaleckis

theory of the financing of investment comes from his principle of increasing risk. Addressing the question of what limits the size of the capital
investment of any firm, Kalecki allowed that a firm which is large relative
to the size of its market would be limited by this, but he held that there is
another factor limiting firm expansion, which is the proportion of the
owners capital that is invested in the firm. For, Kalecki argued, the more of
the wealth of any individual unit of capital that is sunk into one business,
the more at risk is the individuals entire wealth position. This financial
limit on investment means that there is a quantity constraint, as well as a
price of funds constraint, on investment spending.
Keyness ideas about borrowers risk and lenders risk in the General
Theory make a similar point, but Keynes tended to place more emphasis on
the psychological conventions governing the determination of financial
market prices. That is, he argued that speculative activity directed towards
the prices of long-term debt and corporate shares could significantly aect
investment spending. In the face of the radical uncertainty of our knowledge of the future prospects of business, Keynes reasoned that stock
market valuations of the marginal eciency of capital would often represent the results of speculation about the psychology of the market rather
than sensible forecasts of the long-term profitability of corporate capital.
He discussed the ability of central bank policy to move long-term interest
rates to achieve the desired level of investment, and he feared conditions
under which monetary policy would not be able to overcome speculators
liquidity preference suciently. He thus called for the state to take responsibility for ensuring an adequate level of investment through its direct
actions.
James Duesenberry (1958 [1977]) combined the flexible accelerator, in
which the level of investment is explained by the level of output relative to
the level of existing productive capacity, with the financing eects on investment arising from current profit flows and a measure of the existing debt
burden, based on Kaleckis ideas on the determination of investment
finance, into a dynamic marginal eciency of investment and marginal cost
of funds determination of investment. Empirical work in the 1950s and
early 1960s on this approach to investment demonstrated support for
Kaleckis and Keyness ideas.
Theoretical developments of the Post Keynesian theory of investment
were made in the 1970s by Paul Davidson and Hyman Minsky. Davidson
(1972 [1978]) specified the schedule of the demand price of capital goods
as a function of entrepreneurial capitalists subjective rate of discount,
expectations of growth in product demand, their ability to raise the necessary financing, and their calculations of depreciation. His supply price
208
Investment
schedule is given by the size of the existing capital stock and the increasing
cost of production of new capital goods. Davidsons model is thus able to
trace the eects of financial considerations and product market demand
more clearly than Keyness General Theory formulation, which appears to
emphasize only the level of interest rates as the cost of finance, and stock
market prices as the measure of the value of investment projects.
Minskys (1975) formulation takes the cost of production of new capital
to any one firm as given. It then takes the demand price to be the capitalized
value of the expected cash flows from investment, which decreases as the level
of investment rises into the range where use of external financing increases
borrowers risk. In the region of external financing, Minsky depicted increasing lenders risk as a schedule raising supply price at an increasing rate above
the cost of production of new capital. The intersection of the demand and
supply price schedules gives the level of investment spending. Minskys
graphical exposition is somewhat similar to Duesenberrys, but he describes
the details of financial concerns, in terms of both interest rate and debt
burden eects, much more thoroughly and insightfully than anyone else.
Minskys dynamic treatment of the interactions among investment, profits
and debt provides a financial counterpart to Kaleckis portrayal of investment, profits and capital interactions.
In the 1960s, interest in estimating a version of the neoclassical model of
investment determination revived. Neoclassical economists claimed that
profits or other flow measures of the availability of funds only appeared
successful in investment regressions because they were highly correlated
with, and thus were acting as a proxy for, the level of output, which all
theories agreed to be a significant determinant of investment. In the 1980s,
however, work by Steven Fazzari and several dierent co-authors oered
empirical support to Post Keynesian ideas on investment. Fazzari and
Tracy Mott (198687) was the earliest of these to demonstrate support for
the role of output demand, internal finance and debt burden measures in
explaining investment. In later studies, Fazzari and others showed further
the importance of internal financing constraints, following the work of
Kalecki, Minsky and some more recent work based on asymmetric information, in explaining investment.
Most of the theoretical and empirical work discussed above is concerned
with the determinants of business fixed investment. Both Kalecki and
Keynes also wrote about the factors governing inventory investment,
arguing that inventory investment should be influenced by factors similar
to those which determined fixed investment but also be aected by shorterterm movements in the availability of finance and in expectations of sales
relative to current stocks. Fazzaris empirical work on inventory investment
has supported this.
Investment
209
Keynes and Kalecki also both tended to consider the question of the
level of current investment apart from any questions of changes in technique, or capital intensity, which was arguably the main concern of marginal productivity theory. The Cambridge capital critique has questioned
the notion of capital intensity as a measurable concept, and the claim that
investment should be analysed as a process of changes in the ratio of
capital to labour in long-run equilibrium is something that no Post
Keynesian would accept.
Of course, changes in the type of capital must be taken into account in
any long-run analysis of investment. In the Post Keynesian literature these
have been treated mainly under the heading of innovations. Josef Steindls
(1952 [1976]) work on long-run growth within a Kaleckian perspective
argues that in young industries investment is stimulated by the ability of
progressive firms to lower costs through expansion and innovation, and
then to lower prices further in order to drive out higher-cost firms. This
price-cutting maintains capacity utilization at high rates until only a small
group of producers with similar cost structures remain. Price cutting now
oers no advantage to the remaining oligopolists, who thus abandon it.
This in turn decreases the level of investment spending, unless new products or methods of production emerge. In this way Steindl developed his
ideas about absolute concentration into a theory of a long-run tendency
towards macroeconomic stagnation.
Post Keynesians have always acknowledged to some extent the importance of what Keynes called animal spirits as a key influence on the level
of investment spending. Some Post Keynesians have objected that this
makes investment depend too heavily on the subjective reactions of managers to fundamental uncertainty, and therefore underestimates the objective determinants of investment. Kaleckis explanation of investment seems
clearly to rest much more on objective factors, though he did allow that
psychological matters might influence investment activity.
T M
See also:
Capital Theory; Expectations; Innovation; Kaleckian Economics; Keyness General Theory;
Keyness Treatise on Money.
References
Davidson, Paul (1972 [1978]), Money and the Real World, London: Macmillan.
Duesenberry, James (1958 [1977]), Business Cycles and Economic Growth, Westport, CT:
Greenwood Press.
Fazzari, Steven and Tracy Mott (198687), The investment theories of Kalecki and Keynes:
an empirical study of firm data, 19701982, Journal of Post Keynesian Economics, 9 (2),
17187.
210
Investment
Kalecki, Michal- (1990, 1991), Collected Works of Michal- Kalecki, Vols I and II, Oxford:
Clarendon Press.
Keynes, John Maynard (1930 [1971]), A Treatise on Money, London: Macmillan.
Keynes, John Maynard (1936 [1964]), The General Theory of Employment, Interest and Money,
New York: Harcourt Brace.
Minsky, Hyman (1975), John Maynard Keynes, New York: Columbia University Press.
Steindl, Josef (1952 [1976]), Maturity and Stagnation in American Capitalism, New York:
Monthly Review Press.
Joan Robinsons Economics

When Joan Robinson began to study economics in 1922, Marshallian
theory, in the form of the version taught by Pigou, was economics in
Cambridge (CEP I, p. vii; CEP, followed by the Roman number, stands for
J.V. Robinson, Collected Economic Papers, volumes IV, Oxford, Blackwell,
195179. Starred items indicate the 2nd edition). In 192829, she attended
the course Advanced Theory of Value, given by Piero Sraa, who was
calmly committing the sacrilege of pointing out inconsistencies in
Marshall (CEP I, p. vii), and met Richard Kahn, who was preparing his
fellowship dissertation on the Economics of the Short Period; it was the
beginning of a life-long collaboration.
Robinsons first publication, Economics is a Serious Subject. The
Apologia of an Economist to the Mathematician, the Scientist and the Plain
Man, was dedicated to Sraa. By that time the book that was going to give
her fame and academic respectability, The Economics of Imperfect Competition, was finished. Its starting point was Sraas proposal to re-write
the theory of value, starting from the conception of the firm as a monopolist (Robinson 1969, p. 6); its aim was to extend the marginal technique
to all market forms. By this means she hoped to provide an answer to the
challenge posed by Sraa. However, twenty years later she repudiated the
book as a blind alley (Robinson 1978, p. x).
At the same time she was involved in the developments of Keyness new
ideas with the activity of the Cambridge Circus, which met between
January and June 1931, writing two papers on issues being debated there.
In A parable on saving and investment she attacked Keyness argument on
the widows cruse in the Treatise because he was tacitly assuming that
output was unchanged (Robinson 1933, p. 82). In The Theory of Money
and the Analysis of Output she urged Keynes to take the analysis of the
Treatise to its logical conclusion, that is, that output may be in equilibrium
at any number of dierent levels (CEP I, p. 56). Finally, she was one of the
recipients of the first proofs of the General Theory, which she commented
on in June 1935.
Shortly afterwards, she wrote some essays drawing a number of riders
from the General Theory (CEP V, pp. 1856), which were published in 1937
with the title Essays in the Theory of Employment; in the same year she
embarked on the project of writing a version of the General Theory suitable
for teaching to first-year students, which became her Introduction to the
Theory of Employment.
211
212
Joan Robinsons economics
One of the articles collected in the Essays occasioned her encounter with
Michal- Kalecki (CEP V, p. 186). Robinson very soon realized that Kaleckis
analysis was indeed as important as Keyness, and took upon herself the
task of blowing the trumpet for him (ibid.); she later claimed that it was
Kalecki, rather than she herself, who brought imperfect competition in
touch with the theory of employment (Robinson 1969, p. viii).
Kalecki, who had drawn his inspiration from Marxs reproduction
schemes, aroused her interest in them. She began to read Marx in 1940, with
Maurice Dobb as her tutor. Her most substantial work on the subject, An
Essay on Marxian Economics, came out in 1942. The main conclusion of
the book, while revaluating many points of Marxian analysis, was the rejection of Marxs value theory, and over the years she maintained a negative
view of any attempt to solve the problem of transformation (CEP I,
p. 148).
The lesson drawn from the study of Marx in those years was later
summed up by her with the sentence: For me, the main message of Marx
was the need to think in terms of history, not of equilibrium (Robinson
1973, p. x). The influence of Marx appears very clearly in her 1949 review
of Harrods Towards a Dynamic Economics, a book which threw in the
challenge to develop a Keynesian analysis of accumulation in the long run
(CEP II*, p. iii).
The main programme of the 1950s in Cambridge was to develop a longrun analysis of accumulation, that is, to develop an analysis which has
freed itself from the need to assume conditions of static equilibrium (CEP
II*, p. iii). The stumbling-block to the dynamic analysis was given, according to her later recollection, by the lack of an adequate conception of the
rate of profit (CEP II*, p. vi). In fact, on the basis of Keyness and
Kaleckis theory of eective demand, the level of total profits can be determined, while to determine the rate of profit it is necessary to define the
value of the stock of capital, but at the time no one seemed able to do so
(Robinson 1978, p. xvi). She recorded having innumerable discussions with
Piero Sraa but they always consisted in his heading o from errors; he
would never say anything positive. Thus it was not till I found the corn
economy in his Introduction to Ricardos Principles that I saw a gleam of
light on the question of the rate of profit on capital (ibid., p. xvii).
The attempt to extend Keyness short-period analysis to the theory of
long-run development was thus conceived as a return to the classical analysis of accumulation. Her famous books of the late 1950s and early 1960s,
Accumulation of Capital (1956), Exercises in Economic Analysis (1960) and
Essays in the Theory of Economic Growth (1962a), are directed against
models of growth according as they exhibit some kind of inbuilt propensity to maintain full employment over the long run (Robinson 1962a,
Joan Robinsons economics 213

p. 87). The golden-age method, using steady growth models with full
employment, was provided to examine the relation between accumulation
and the rate of profit (CEP V, p. 21). The dierence between the equilibrium method and the historical method was seen as a dierent treatment
of time: To make a comparison between two situations, each with its own
future and its own past, is not the same thing as to trace a movement from
one to the other (Robinson 1960, p. v).
The long struggle to escape (CEP III*, p. 52) from a conception in
which accumulation is seen as a substitution of labour for capital in a given
state of technical knowledge meant reinstating the possibility of the analysis of innovations and technical progress, as Smith, Ricardo and Marx
had done.
In her attempt to analyse the relationship between the rate of profit and
the choice of techniques, Robinson was faced with the question of the
meaning to be given to the expression quantity of capital. In her 195354
article on the production function, and then in the Accumulation of Capital,
she had invented a pseudo-production function, as Robert Solow later
called it (CEP V, p. 82), in order to be able to list the techniques specified
in a supposed book of blueprints, which represented the state of technical knowledge at a given point of time. The pseudo-production function
was meant to show the possible equilibrium positions corresponding to
dierent values of the rate of profit. So she encountered the phenomenon
of reswitching, namely that:
[O]ver certain ranges of a pseudo-production function the technique that
becomes eligible at a higher rate of profit (with a correspondingly lower realwage rate) may be less labour intensive (that is, may have a higher output per
man employed) than that chosen at a higher wage rate, contrary to the rule of a
well-behaved production function in which a lower wage rate is always associated with a more intensive technique. (CEP IV, pp. 1445)
With the publication in 1960 of Sraas Production of Commodities by

Means of Commodities the basic tenets of his criticism of neoclassical
theory could be seen more clearly. Sraas message has a twofold significance, according to Robinson: to knock out the marginal productivity
theory and to re-establish the classical doctrine that the rate of profit on
capital depends upon the technical structure of production and the share
of wages in net output (CEP V, p. 95).
The conviction that it is possible to keep the scientific and ideological
levels of analysis separate is at the core of Robinsons attitude to economics. In 1962 she presented her methodological ideas in Economic Philosophy, where she argued that in scientific discourse it is possible to distinguish
empirical propositions from metaphysical propositions, as Karl Popper
214
Joan Robinsons economics
had maintained (Robinson 1962b, p. 3). Unfortunately, as she commented

in her Exercises in Economic Analysis: Economics does not oer, like the
well-developed natural sciences, a body of knowledge which the lay public
can accept as established (Robinson 1960, p. xv).
When she became Professor of Economics at Cambridge in 1965, she
chose as the topic of her inaugural lecture, The new mercantilism, a
denunciation of the mystique of free trade in historical practice and in the
theoretical tradition since the time of Adam Smith (CEP IV, p. 4). She
argued that contemporary neo-mercantilist philosophies and policies are
always followed and theorized when the benefits of free trade are in danger
(CEP IV, pp. 1213).
In the early 1970s Robinson came insistently to the question of identifying the pars construens of her thought in the classical tradition, revived by
Sraa, which flows from Ricardo through Marx, diluted by Marshall and
enriched by the analysis of eective demand of Keynes and Kalecki
(Robinson 1973, p. xii). However, in the work of reconstruction she found
herself in disagreement with some of her allies in the battle against neoclassical economics. One point in particular became central in the discussion,
that is, the maintenance of a concept of a long-run equilibrium in the
context of historical analysis (Robinson 1980, p. 128).
It is the criticism of the concept of equilibrium, not only of neoclassical
equilibrium, which she sees as the legacy of Keynes; therefore the main
instrument with which to attack the neoclassical theory should be the distinction between historical time and logical time. In this respect she found
Sraas language in Production of Commodities limited, because what it is
oered is a purely logical structure an elaborate thought experiment.
There is no causation and no change (Robinson 1980, p. 132). This is why
she sees it as more promising to begin again with Keynes, who discusses
events in terms of processes taking place in actual history (Robinson 1979,
p. xiv).
At the end of her life Robinson became increasingly dissatisfied with economics and more and more disillusioned with it as a body of knowledge
which could be used to solve problems in the real world. She was increasingly concerned with those fundamental issues which are obscured rather
than clarified by contemporary economic theory. Her last paper, published
posthumously, originally had a telling title, Spring cleaning: We should
throw out all self-contradictory propositions, unmeasurable quantities and
indefinable concepts and reconstruct a logical basis for analysis with what,
if anything, remains (Robinson 1985, p. 160).
This is the legacy that Robinson has handed down to us. (See Marcuzzo
1996, 2001.)
M C M
Journal of Post Keynesian Economics
215
See also:
Cambridge Economic Tradition; Capital Theory; Growth and Income Distribution; Growth
Theory; Kaleckian Economics; Sraan Economics; Time in Economic Theory.
References
Marcuzzo, M.C. (1996), The writings of Joan Robinson, in M.C. Marcuzzo, L.L. Pasinetti
and A. Roncaglia (eds), The Economics of Joan Robinson, London: Routledge, pp. 33063.
Marcuzzo, M.C. (2001), Joan Robinson: une qute passionne de la rationalit, in G.
Harcourt (ed.), Lconomie rebelle de Joan Robinson, Paris: LHarmattan, pp. 2758.
Robinson, J. (1933), A parable on saving and investment, Economica, 13 (39), 7584.
Robinson, J. (1956), The Accumulation of Capital, London: Macmillan.
Robinson, J. (1960), Exercises in Economic Analysis, London: Macmillan.
Robinson, J. (1962a), Essays in the Theory of Economic Growth, London: Macmillan.
Robinson, J. (1962b), Economic Philosophy, London: Watts.
Robinson, J. (1969), The Economics of Imperfect Competition, London: Macmillan, second
edition.
Robinson, J. (1973), Preface to J.A. Kregel, The Reconstruction of Political Economy: An
Introduction to Post-Keynesian Economics, London: Macmillan, pp. ixxiii.
Robinson, J. (1978), Contributions to Modern Economics, Oxford: Blackwell.
Robinson, J. (1979), The Generalization of the General Theory and Other Essays, London:
Macmillan.
Robinson, J. (1980), Further Contributions to Modern Economics, Oxford: Blackwell.
Robinson, J. (1985), The theory of normal prices and reconstruction of economic theory, in
G.R. Feiwel (ed.), The Theory of Normal Prices and Reconstruction of Economic Theory,
London: Macmillan, pp. 15765.

Political economy experienced a revival in the 1960s in many nations of
the world, leading to the inception of many associations with journals
to support a growing number of adherents and fellow travellers. Post
Keynesians never developed their own formal (multinational) association,
and it was not until the late 1970s that specifically Post Keynesian journals
emerged, starting in the UK with the Cambridge Journal of Economics in
1977. In the same year, the father of Post Keynesian economics in the
United States, Sidney Weintraub (191483), along with a former student,
Paul Davidson, sent out invitations to potential subscribers to another new
journal. To their shock and amazement, they received cheques from
more than 400 subscribers within a month of the mailout (see Davidson,
JPKE, Fall 1998, p. 3). The first issue of the Journal of Post Keynesian
Economics (JPKE) thus emerged in the Fall of 1978, published by M.E.
Sharpe of New York, with a pre-eminent international Honorary Board of
Editors, and the editors Davidson and Weintraub overseeing the journal
through the usual four issues a year.
The first issue of the journal included an editorial Statement of
Purposes (JPKE, Fall 1978, pp. 37), which made it clear that the journal
216
was to be concerned with innovative theoretical work that can shed fresh
light on contemporary economic problems while contest[ing the] orthodoxy that dominates journals in the US. The editors believed that
Innovative ideas on inflation and unemployment have been routinely suppressed by prominent journals. Such ideas have tried to explain the real
world as well as provide a reliable guide to public policy. They cite some
of the greats from the distant past Adam Smith, David Ricardo, Karl
Marx, J.S. Mill, W.S. Jevons, Alfred Marshall and J.M. Keynes as well as
some from the late 1970s Joan Robinson, Nicholas Kaldor, Richard Kahn,
Michal- Kalecki, Abba Lerner, J.K. Galbraith and Hyman Minsky as
forging the central intellectual spirit of Post Keynesian economics. This
spirit incorporates a monetary theory of production, where financial relationships influence economic processes in the short and long runs (money
matters), due to hysteresis, path dependency and fundamental uncertainty.
Money matters because the financial system generates credit for productive
and financial activities, and the holding of money and credit influences
velocity, money supply and thus GDP. Special reference is given in the
journal to the problems of uncertainty, credit and demand in aecting inflation, unemployment, corporate power, capitallabour relations, demand
management tools and strategies to enhance the general welfare in the elemental, benign sense of that elliptical concept. Significantly, the editors
added that: It is not a new sect that we seek to foster; it is instead a reasoned
debate with a fair shake for innovative, unorthodox attitudes. The term
post Keynesian [sic] will thus be broadly interpreted, spotlighting new
problems and revealing new theoretical perspectives.
Apart from general articles, the journal has included many interesting
symposia, comments, book reviews, an editorial corner and a series of lively
and humanistic academic biographies (written mainly by board member
Georey Harcourt). Since the death of Weintraub in 1983, Paul Davidson
has been the sole editor and main force behind the journal. His views have
some bearing on its content and trend. For instance, he has never been too
impressed by Sraan themes, and has sought to dierentiate Keyness
message from that of Kalecki (although both have impacted on the JPKE).
Through its 25-year history the journal has concentrated on relating the
core theory of Post Keynesian economics concerning money, uncertainty
and demand to new developments, trends or problems in the world. For
instance, in the 1970s and 1980s it paid special attention to the problems
associated with stagflation, while during the 1990s and early 2000s it gave
more attention to financial crises, global imbalances and conflicts.
The principal themes of the core theory encouraged by Davidson seem to
be, wittingly or unwittingly, a symbolic reflection of the composition of the
Honorary Board of Editors. A core theme running through the journal, as
217
Philip Arestis says (ironically quoting Michal- Kalecki (18991970)), is that

Post Keynesian analysis firmly embraces the view that the institutional
framework of a social system is a basic element of its economic dynamics
(JPKE, Summer 1989, p. 611). This reflects the concerns of (especially)
board members Arestis, Galbraith (chair), Gunnar Myrdal (18981987),
Daniel Fusfeld, Robert Heilbroner, Hyman Minsky (191996), Wallace
Peterson and Warren Samuels. Of prime importance in this respect is an
understanding of capitalism as a system, comprising a mixed economy
state and corporation; a heterogeneous series of social classes workers,
capitalists and salaried professionals; a corporate system of big and small
firms; a complex system of finance including banks, institutional investors
and central authority; and a series of nation-states, global institutions and
networks.
The main institutional themes of the journal link, directly or indirectly,
to the theory of circular and cumulative causation (CCC), emanating especially from board members Myrdal, Nicholas Kaldor (190886) and A.P.
Thirlwall. According to CCC, the main institutions and sectors of the economic system link together in a complex circuit of cybernetic feedback and
interaction. Supply and demand are interdependent. For instance, households are not only consumers but also investors in durable structures. This
requires a detailed analysis of habits, social conventions, bounded rationality, and a hierarchy of needs from basics to luxuries (including Veblen goods
and conspicuous consumption). Investment demand is associated with
economies of scale/scope, new technology, complex dynamics, changes in
capacity utilization and structural change. A proper system of government
spending enhances infrastructure, knowledge and organization. And the
balance of payments constraint recognizes the impact of import elasticities, world income and non-price competition on economic growth.
Demand is the prime mover of the CCC circuit, since it links to the creation
of new needs and changes to the systems of production and distribution,
and hence links productivity and exports through economies of scale, learning by doing and structural modifications to needs and technologies. (See
the debate on Eective demand, JPKE, Spring 2001, pp. 375440; and the
Symposium on Thirlwalls Law and the BOPC, JPKE, Spring 1997, pp.
31186.)
A primary source of change in a Post Keynesian world that is copiously
developed in the journal is fundamental uncertainty, or non-ergodic dynamics, a theme expounded by Davidson as well as JPKE board members
Donald Katzner, Minsky, G.L.S. Shackle (190392) and Douglas Vickers.
In their view, investment is largely aected by the prevailing business climate
and demand, especially the degree of confidence in the future, and hence
expected profits, la chapter 12 of Keyness General Theory of Employment,
218
Interest and Money (1936). The future, however, is unknown and therefore
uncertain (rather than strictly probabilistic): the greater the level of uncertainty, the lower the expected rate of profit and rate of investment. Firms
engage in routines, organizations and institutions, such as accounting
notions of cash flow, net worth and mark-up pricing, to provide stability
and structured activity in a world of uncertainty. This enables firms to invest
in capital goods and consumer goods production based largely on credit
with some degree of confidence that profits will flow from such activities.
But, despite all this, fundamental uncertainty and ignorance about the
future still prevail, which periodically lead to booms and recessions of
varying magnitudes. (See the Symposium on Investment, JPKE, Summer
1992, pp. 42396; the issue mainly on Uncertainty, Fall 1993, pp. 354; and
the Spring 1996 issue.)
Editor Davidson has published a lot on demand and supply equations
and conditions that are interdependent. The supply side links to work on
non-price competition, degree of monopoly, mark-up pricing principles
and the megacorp, especially by board members Alfred Eichner (193788),
Fred Lee, J. Barkley Rosser Jr, Malcolm Sawyer and Nina Shapiro.
National income, equating price level (p) times output (q), from the supply
side (Ys), equals:
Ys pq(w/A)q.
Weintraub explained inflation through a wagecost mark-up equation, p
w/A, where is the average mark-up of prices over unit wage (variable)
costs, w/A, w the average money-wage rate and A the average product per
worker. The mark-up by firms, , is said to be remarkably stable, dependent
mainly on the relationship between wages and labour productivity (JPKE,
Winter 198182, pp. 291300). Inflation has thus been linked in the JPKE
to wages upwardly deviating from productivity, leading firms to adjust their
prices accordingly.
National income in nominal terms, from the demand side (Yd), includes
workers wages (W) and capitalists profit (), which equals consumption
(C) plus investment (I) plus the government budget deficit (GD) and the
trade surplus (TS):
Yd WCIGDTS.
In Kaleckian models, workers spend all of their income, and capitalists
spend what they receive (adjustments have been made for workers and capitalists saving some of their income and so on). Under these conditions,
aggregate profit equals the consumption of capitalists plus investment plus
219
the budget deficit plus the trade surplus. Workers press for wage claims in
the industrial relations arena, while firms seek claims through their target
price mark-ups (JPKE, Fall 1991, pp. 93110). The JPKE has published
papers supporting the conflicting claims theory, where inflation is due to
aggregate nominal income claims wages and profits exceeding the total
available income.
The JPKE specializes in developing endogenous explanations for economic phenomena, such as the creation of credit, business cycles, financial
crises and exchange rates. This material reveals that during business cycle
upswings such as 198487 or 19962000 (199297 in parts of Asia) the
generation of euphoria leads to a high rate of investment, as well as a stock
market boom, financed largely by endogenous credit. According to Hyman
Minskys distance memory hypothesis, firms typically forget about previous financial crises and recessions and get caught up in the euphoric environment. Traders speculated in the late 1990s2000 period, for instance,
about new rules to the game, and technology stocks not needing fundamentals in the new business environment, which supposedly justified
further credit expansion. The journal has explored endogenous money and
credit responding to higher demand through instruments such as bank
bills, certificates of deposit, capital inflow, financial innovations and reserve
bank accommodation. Forces endogenous to the upswing such as higher
oil and energy prices, interest rate pressures, speculative bubbles and heightened global conflict bring about a collapse of prospective yields, greater
uncertainty, a speculative bubble crash, a declining rate of investment
demand and further financial crises, for instance during 20002002 as
euphoria turned to global and national pessimism and recession (JPKE,
Spring 1990 and Winter 20002001).
This leads the journal to concentrate on policies that moderate the
instabilities of the business cycle, such as incomes policies, global institutions, basic income schemes, prudential financial policies and organizational arrangements (see JPKE Winter 199798, Spring 2000 and Summer
2000). A recurring global policy proposal is monetary reform, where the
onus is on (current account) surplus nations increasing their eective
demand, thereby restoring some degree of world balance. A cooperative
global approach to balance of payments problems is seen to be better than
one where deficit nations bear the brunt of adjustment through fiscal and
monetary deflation. The cooperative policy could thus help to moderate
various conflicts, including terrorism, war, crime and industrial instability,
through the provision of global stability, trust, agreement and income.
Over the past quarter of a century the Journal of Post Keynesian
Economics has been instrumental in promoting theoretical and policy
insights that enhance the democratic and participatory workings of the
220
economy in the pursuit of full employment and price stability. It has

tracked and critically analysed the contemporary economic performance of
the US and world economies. The journal has done well to propagate a
coherent alternative analysis of economic theory and policy in a dicult
era of neoliberalism. Advances in knowledge have been made in relation
to a monetary theory of production, set in an environment of fundamental uncertainty, circular and cumulative causation, and interdependencies
between supply and demand. The editor has encouraged the building of
empirical evidence, the linking of theory and practice, and the development
of policy prescriptions that are innovative yet relatively pragmatic, but
always inspired by the need for fundamental reforms of capitalism. The
ability of the journal to withstand editorial, systemic and disciplinary
instabilities and progress further will influence the future of Post Keynesian
economics in the United States and elsewhere.
P A OH
Kaldorian Economics
Nicholas (Miklos) Kaldor (190886) was one of the most original and controversial economists of the twentieth century. The Economist newspaper
once described him as the best known economist in the world not to have
received the Nobel Prize. In the 1930s and 1940s he made fundamental
contributions to the theory of the firm; welfare economics; trade cycle
theory; capital theory and Keynesian economics. In the 1950s, he turned his
fertile mind to public finance and to growth and distribution theory, and
was the joint architect with Joan Robinson and Richard Kahn of the Post
Keynesian school of economics which extended Keynesian modes of
thinking to the long run. Then in the 1960s he turned his attention to the
applied economics of growth and initiated an enormous secondary literature related to the idea of manufacturing industry as the engine of growth
based on static and dynamic increasing returns to scale. In the 1970s he led
world-wide the assault on the doctrine of monetarism which, as he
described it, spread with the virulence of a plague from North America
under the influence of Milton Friedman to infect academic thinking and
policy making in several parts of the world, including most notably the
United Kingdom during the government of Margaret Thatcher in the
1980s. He lost the battle but won the war because monetarism as a coherent intellectual doctrine is now dead.
It is clear from the above that Kaldor led several lives as an economist;
his range of interests was wide, but he also had a vision of how capitalist
economies function and a strong intuition concerning what is important
and what is unimportant, what is cause and what is eect. There is a
Kaldorian economics and an interesting story to tell.
Kaldor was Hungarian by birth. As a boy he attended the famous Minta
school in Budapest, and then at the age of 17 attended the Humboldt
University in Berlin to study economics for eighteen months before coming
to the London School of Economics (LSE) in 1927, where he fell under the
influence of Allyn Young (who tragically died in 1929, aged 53) and then of
Lionel Robbins and (later) Friedrich von Hayek. In 1930 he graduated with
first-class honours and stayed on at the LSE first as a research assistant and
then as an assistant lecturer. The Keynesian revolution was still six years
o, and his early research work was in the Austrian tradition an analysis,
for example, of the overcommitment of Austrian industry and the problem
of the Danubian states. It was not long, however, before Kaldor crossed
swords with Robbins and Hayek and became one of the first converts at the
221
222
Kaldorian economics
LSE to the thinking in Keyness General Theory, along with Abba Lerner
and Ursula Hicks. While Kaldor disagreed with some of the details of the
General Theory, and made important contributions himself to its understanding, he never wavered from the thrust of its central message that monetary production economies are fundamentally dierent from barter
economies, and that unemployment can exist for long periods of time even
in the presence of wage and price flexibility, because of uncertainty associated with the peculiar properties of money.
Kaldors first major theoretical contributions came in 193435 with four
papers: A classificatory note on the determinateness of equilibrium, in
which he was the first to coin the term cob-web theorem to describe oscillations around an equilibrium; The equilibrium of the firm; Mrs.
Robinsons Economics of Imperfect Competition; and Market imperfections and excess capacity. In the next five years, including the first year of
the war, there appeared a further spate of papers in diverse fields (Thirlwall
1987 and Targetti 1992 contain a full bibliography). There was his major
survey of capital theory; his attack on Arthur Pigous theory of how wage
cuts aect unemployment it must be through a reduction in the rate of
interest; his critique of Edward Chamberlin and the distinction between
monopolistic and imperfect competition; his debate with Hayek over
capital intensity and the trade cycle; his introduction of compensation tests
into welfare economies; his classic paper Speculation and economic stability which John Hicks described in personal correspondence as the culmination of the Keynesian revolution in theory you ought to have got
more honour for it (Thirlwall 1987, p. 75, n. 46); and his 1940 nonlinear
model of the trade cycle.
During the Second World War the LSE was evacuated to Cambridge,
and Kaldor became more acquainted with the Cambridge economists, particularly Joan Robinson, Richard Kahn, Piero Sraa, Pigou and Keynes
himself. He spent most of the war years working on aspects of public
policy, both national and international, related both to the war and preparations for the peace. In particular, he played a major role in the analysis
and thinking behind the two Beveridge reports on Social Insurance (in
1942) and Full Employment in a Free Society (in 1944), and also the construction of national income accounts, then in their infancy. He emerged
from the war with a high reputation as an incisive applied economist, which
led to his appointment on several international commissions, and then in
1947 as research director of the Economic Commission for Europe in
Geneva headed by Gunnar Myrdal. In 1949 he returned to Cambridge as
a Fellow of Kings College and a member of the Economics Faculty, where
he remained for the rest of his life.
His interest in public finance, and particularly tax matters, deepened
Kaldorian economics 223

when he was appointed in 1951, along with John Hicks, to the Royal
Commission on the Taxation of Profits and Income. Kaldor was the author
of a trenchant Minority Report and also a classic book entitled An
Expenditure Tax arguing the case for taxing expenditure rather than
income, both on grounds of equity and as a means of promoting growth.
His tax expertise was later sought in several developing countries, and led
to his appointment in the United Kingdom as a special tax adviser to the
Chancellor of the Exchequer in two Labour governments (196469 and
197476).
Despite his multifarious contributions to economic theory and policy,
Kaldor will be remembered in the history of economic thought largely for
his work in growth economics and his challenge to equilibrium theory. In
the mid-1950s and early 1960s there was his pioneering work on growth and
distribution theory. In the mid-1960s there was his innovative thinking on
the applied economics of growth, and then from the 1970s to his death
there was his constant assault on the assumptions, predictions and usefulness of the Walrasian, general equilibrium framework of analysis for an
understanding of the dynamics of capitalism in the real world. What are
the mainsprings of growth in developed and developing countries? Does it
make sense to separate capital accumulation and technical progress as in
mainstream neoclassical growth theory? Why does the capitaloutput ratio
remain roughly constant despite an ever-increasing ratio of capital to
labour? Why does fast growth seem to be associated with industrialization?
What determines the growth of industry in a closed economy (including the
world) and in an open economy? Why is there a tendency for levels of development between regions and countries to become polarized, contrary to
the predictions of neoclassical growth theory? These are the major questions that Kaldor attempted to answer in a series of profound and provocative papers over 30 years (for example, Kaldor 1956, 1957, 1961, 1966,
1970, 1972, 1996).
In 1956 Kaldor revolutionized the theory of the functional distribution
of income by showing that the share of profits in national income must be
related to the share of investment in national income and the propensity to
save out of wages and profits, and that the orthodox neoclassical theory of
distribution based on relative factor prices and quantities is not only theoretically fraught with problems, but, in any case, unnecessary. The Kaldor
theory of distribution is beautiful in its simplicity. Let full-employment
income (Y) be divided between consumption (C) and investment (I), with
consumption out of wages equal to cwW and consumption out of profits
equal to cpP (where W is wages, P is profits and cw and cp are the propensities to consume out of wages and profits, respectively). Therefore, Y
cwWcpPI. But PYW. Therefore PcpPIswW, where sw is the
224
Kaldorian economics
propensity to save out of wages. The share of profits in income is therefore

equal to P/Y(I/Y)/(sp sw)sw/ (sp sw), where sp is the propensity to save
out of profits. Profits must be the dependent variable and investment the
independent variable because capitalists can decide what to invest but they
cannot decide what they earn.
Kaldors 1957 and 1961 growth models introduce the innovation of the
technical progress function (TPF) to replace the neoclassical production
function, which makes what is an artificial distinction between movements
along a function and shifts in the whole function. Kaldor was adamant that
capital accumulation and technical progress go together; most technical
progress requires capital accumulation for its embodiment and there is
unlikely to be much new capital accumulation without technical progress.
The TPF relates the rate of growth of output per worker (r) to the rate of
growth of capital per worker (k). In linear form: rab (k). The position
of the function (a) depends on autonomous productivity growth (for
example, learning by doing) and the slope of the function (b) depends on the
technical dynamism of the economy. Equilibrium growth at a constant
capitaloutput ratio is given by ra/(1b). Countries grow at dierent
rates with the same capitaloutput ratio because of dierences in the parameters of the technical progress function. The TPF is important not only
because it provides a more realistic representation of the growth process, but
also because it provides an explanation of why the capitaloutput ratio is no
higher in rich countries than in poor countries, contrary to the prediction of
orthodox neoclassical growth theory; countries are simply on dierent
TPFs. Kaldors TPF is thus the precursor, and true progenitor, of new
growth theory (or endogenous growth theory), which also seeks to explain
why the marginal product of capital apparently does not fall as countries get
richer and invest more. There are technological forces that keep the
capitaloutput ratio from rising, such as human capital formation, research
and development and technological spillovers from trade.
Kaldor was not made a professor at Cambridge until 1966, aged 58, but
in his famous Inaugural Lecture (Kaldor 1966) he turned his attention to
the applied economics of growth and presented a series of growth laws
which have subsequently been widely tested in dierent contexts. Kaldors
thesis was that manufacturing industry is the engine of growth for two
major reasons. First, manufacturing industry is subject to increasing
returns which other sectors are not (at least to the same extent, and certainly not agriculture). This hypothesis is also known as Verdoorns Law
(see McCombie et al. 2002), which Kaldor revived, it having lain eectively
dormant for 17 years. Second, manufacturing industry generates fast
growth because it draws resources in from sectors where the marginal
product is less than the average, so that productivity growth is induced

outside of the manufacturing sector. Today, the evidence for Kaldors laws
is most clearly seen in the newly industrializing countries of South-East
Asia,where GDP growth is rapid and the share of manufacturing industry
in GDP is also rising fast.
But what determines the growth of industry? In the open economy it is
the growth of exports in a circular and cumulative process (Kaldor 1970),
although subject to a balance of payments constraint. In the closed
economy, such as the world economy, it is land-saving innovations in agriculture that are crucial for industrial growth (and the performance of developed countries) as an oset to diminishing returns. In his 1986 Hicks
Lecture on Limits to growth (and elsewhere, for example, Kaldor 1996),
he emphasizes the importance of an equilibrium terms of trade between the
two sectors of agriculture and industry (developing and developed countries) for maximizing the growth rate of the economy as a whole.
In all these fields of theoretical and empirical enquiry mentioned above,
Kaldor attracted many disciples, but also many adversaries. Academic hostility came from asking awkward questions about neoclassical economics
in general and equilibrium theory in particular, and was understandably
most prevalent in the US. Kaldors legacy to the profession is nearly 200
articles, pamphlets and books, many of the former being collected in his
nine volumes of Collected Essays. Although he wrote no grand treatise,
these volumes will provide a lasting testimony and monument to the energy,
creativity and endeavour of one of the greatest economists of the twentieth century.
A.P. T
See also:
Balance-of-payments-constrained Economic Growth; Cambridge Economic Tradition;
Equilibrium and Non-equilibrium; Growth and Income Distribution; Growth Theory;
Income Distribution.
References
Kaldor, N. (1956), Alternative theories of distribution, Review of Economic Studies, 23 (2),
83100.
Kaldor, N. (1957), A model of economic growth, Economic Journal, 67 (268), 591624.
Kaldor, N. (1961), Capital accumulation and economic growth, in F. Lutz (ed.), The Theory
of Capital, London: Macmillan, pp. 177222.
Kaldor, N. (1966), Causes of the Slow Rate of Economic Growth of the United Kingdom,
Kaldor, N. (1970), The case for regional policies, Scottish Journal of Political Economy, 17
(3), 33748.
Kaldor, N. (1972), The irrelevance of equilibrium economics, Economic Journal, 82 (328),
123755.
Kaldor, N. (1996), Causes of Growth and Stagnation in the World Economy (Mattioli Lectures),
226
Kaleckian economics
McCombie, J., M. Pugno and B. Soro (2002), Productivity Growth and Economic Performance:
Essays on Verdoorns Law, London: Palgrave.
Targetti, F. (1992), Nicholas Kaldor, Oxford: Clarendon Press.
Thirlwall, A.P. (1987), Nicholas Kaldor, Brighton: Harvester Wheatsheaf.
Kaleckian Economics
Kaleckian economics may be broadly defined as the economic theories enunciated by Michal- Kalecki (18991970), and the extensions of those theories
by economists who were influenced by him. Kalecki was a Polish engineer
and mathematician who taught himself economics in a left-wing political
milieu during the 1920s, where the main intellectual influences were AustroMarxism, Rosa Luxemburg, Mikhail Tugan-Baranovsky and Henryk
Grossmann. From 1929 to 1936 he was employed at the Business Cycle and
Prices Institute in Warsaw, where the first national income statistics for
Poland were constructed. In 1933 he published his first analysis of the business cycle under capitalism, arguing that it was due to the instability of
investment, which in turn was caused by fluctuations in capitalists profits.
Investment was crucial. Under capitalism, in Kaleckis view, investment is the
main determinant of aggregate demand as well as, in its turn, determining
profits, where capitalists costs are mainly accounted for by wages, which are
by and large consumed. This was summarized in an aphoristic precis of
Kaleckis theory (attributed by some to Joan Robinson, and by others to
Nicholas Kaldor): Workers spend what they earn, capitalists earn what they
spend. This can easily be derived from the well-known Keynesian savings
identity, in which saving (S) is by definition equal to gross investment (I), plus
the fiscal deficit (government expenditure minus taxation, GT), plus the
trade surplus (exports minus imports, XM). Assume that there are only
two classes in society, capitalists and workers, earning profits and wages
respectively, which can be saved or consumed. The saving identity therefore
represents workers and capitalists saving (Sw and Sc):
SI(GT)(XM)Sw Sc.
If workers saving is deducted from both sides of the equation, then the
saving identity shows only capitalists saving:
Sc I(GT)(XM)Sw.
Since profits can only be saved or consumed, adding capitalists consumption (Cc) to the right-hand side of the equation gives an expression for
profits (P):
PCc Sc I(GT)(XM)Cc Sw.
Kaleckian economics
227
Profits are therefore equal to gross investment, plus the fiscal deficit, plus
the trade surplus, plus capitalists consumption, minus workers saving. The
greater is capitalists expenditure on investment or their own consumption,
or the fiscal surplus, or the expenditure of foreign residents on exports, the
greater will profits be. Higher profits will tend to result in higher investment
until excess capacity emerges and investment is reduced, causing profits to
fall and a decline in economic activity to continue until excess capacity is
eliminated and investment starts to rise. Higher profits then finance higher
investment and stimulate a boom in economic activity.
In 1936 Kalecki left Poland for Stockholm and eventually for London,
where Joan Robinson recruited him to Keyness circle. While critical of
Keyness equilibrium reasoning, he readily participated in Cambridge and
later Oxford discussions on the possibilities of full employment under capitalism. At this time he developed his pricing analysis, in which the markup over prime costs is determined by imperfect competition, and an
analysis of corporate finance in which external finance is a liability that
enhances financial risks, as well as providing liquidity. After the Second
World War, Kalecki worked for nearly ten years for the United Nations,
where he studied in detail the problems of developing countries. Out of this,
in later years, came an analysis of economic development focusing on
financial bottlenecks to capital accumulation in the developing countries,
in a context of socio-economic structural obstacles, poverty, rural backwardness and food supply, to capitalist primary accumulation. In 1955
Kalecki returned to Poland. In the dislocation caused by Stalinist overinvestment, he emphasized the limited eectiveness of investment because of
the need to maintain adequate levels of consumption and avoid excessive
imports. He was a strong critic of market socialism, arguing that market
mechanisms are less ecient than an eectively adjusted and centralized
investment programme. Kalecki and his associates were subjected to political attacks and an anti-Semitic purge in 1968.
In the years after Keyness death, Joan Robinson championed Kaleckis
work for its radical criticism of capitalism, namely that capitalism is unstable (the business cycle), tends to regressive distributional values (costminimization holds down wages, while high profits are necessary to
maintain investment), and is hostile to full employment (because it undermines labour discipline) (Robinson 1966). Kaleckis exposition of his analysis in the form of mathematical models based on national income
identities made his work attractive to the first generation of Keynesian
model-builders, in particular Lawrence Klein and David Worswick. They
were attracted by models which gave a more systematic account of business
cycles than Keyness (Klein 1947).
During the 1950s Kalecki was influential in the monopoly capitalism
228
Kaleckian economics
school of Marxists, through the work of Paul Sweezy and Josef Steindl.
Kaleckis analysis shows how the problem of realizing surplus value as
profits, in twentieth-century capitalism, was alleviated by corporate investment and deficit spending by governments. At the same time, the absence
of competition gives capitalists monopoly profits, which make excess
capacity more tolerable. Such excess capacity in turn reduces the capitalists inclination to invest, causing a tendency to economic stagnation
(Baran and Sweezy 1966; Steindl 1952).
Kaleckis ideas were at the forefront of the emergence of Post Keynesian
economics during the 1970s. Here Kaleckian economics provided a clear
and consistent alternative to the neoclassical synthesis of Keynesian ideas
with Walrasian general equilibrium. Kalecki could provide not only a
theory of the business cycle (an essential element of any economic analysis
after the return of economic instability to capitalism in the 1970s), but also
microeconomic foundations, which are largely absent in Keyness General
Theory. (Their absence had facilitated the neoclassical and monetarist
interpretation of Keynesian unemployment as being due to wage inflexibility.) Kalecki provided a more radical microeconomic explanation, in terms
of monopoly and excess capacity reducing the propensity to invest out of
profits (King 1996; Sawyer 1985). In this way Post Keynesian analysis
spliced Kaleckis price and business cycle theory on to more orthodox
Keynesian concerns about aggregate demand and full employment.
However, Post Keynesians have, by and large, preferred to overlook two
aspects in which the work of Kalecki and Keynes is less than compatible.
The first of these arises out of their respective treatment of expectations
and uncertainty. Expectations play a central role in Keyness explanation of
the instability of investment, to which both theorists attributed the business
cycle. In Kaleckis view, business confidence is largely determined by
current profits, so that further analysis of the subjective elements entering
into businesspeoples expectations is unnecessary. Uncertainty then plays a
crucial role in Keyness liquidity preference theory of money. Coming from
outside the Marshallian tradition, Kalecki did not find it necessary to postulate any aggregate demand for and supply of money, outside the wholesale money markets, and he took it to be a central feature of capitalism that
the banking system accommodates business demand for credit. Money is
therefore endogenous to the system, and uncertainty is less important in
portfolio demand for money (Keyness speculative demand) than changes
in short-term interest rates, relative to the long-term rate of interest.
Kalecki was also critical of Keyness emphasis on the long-term rate of
interest (the yield on long-term bonds) as a determinant of investment. That
rate of interest was shown to be relatively stable, and therefore was of little
use in explaining the instability of investment. Keynes resolved this problem
Keyness General Theory
229
by arguing that the expected return on investment that is in excess of the

long-term rate of interest (his marginal eciency of capital) is volatile, and
therefore accounts for the instability of investment. Kalecki argued instead
that investment is volatile because the internal liquidity of the corporate
sector that is free of external financial liabilities (and is therefore available
for investment without imposing potentially ruinous financial overheads on
companies) fluctuates with profits and the degree of external financing. This
is Kaleckis principle of increasing risk. W.H. Locke Anderson made a pioneering study of this in the early 1960s. The chief exponent of Post
Keynesianism as a theory of finance capital, Hyman Minsky, used Kaleckis
theory of the business cycle, but developed his own analysis of investment
financing based on Keynesian expectations and Irving Fishers debt deflation theory of economic depressions (Minsky 1986).
While Post Keynesians have tended to use Kaleckis analysis selectively
to fill the lacunae in Keyness economics, and the collapse of Communism
has seriously limited the interest in Kaleckis economics of socialism, recurrent economic crises in developing and newly-industrialized countries, and
volatile financial conditions in the older capitalist countries, oer scope for
new developments in, and applications of, Kaleckis economics.
J T
See also:
Business Cycles; Development Finance; Expectations; Financial Instability Hypothesis;
Pricing and Prices; Profits; Socialism; Uncertainty; Underconsumption.
References
Baran, P. A. and P.M. Sweezy (1966), Monopoly Capital: An Essay on the American Economic
and Social Order, New York: Monthly Review Press.
Kalecki, M. (199097) Collected Works of Michal- Kalecki, edited by Jerzy Osiatynski, Oxford:
Clarendon Press.
King, J.E. (ed.) (1996), An Alternative Macroeconomic Theory: The Kaleckian Model and PostKeynesian Economics, Boston: Kluwer.
Klein, L.R. (1947), The Keynesian Revolution, New York: Macmillan.
Robinson, J. (1966), Kalecki and Keynes, in Economic Dynamics and Planning: Essays in
Honour of Michal- Kalecki, Oxford: Pergamon Press, pp. 33541.
Sawyer, M.C. (1985) The Economics of Michal- Kalecki, London: Macmillan.
Steindl, J. (1952), Maturity and Stagnation in American Capitalism, Oxford: Blackwell.

Most economists recognize that Keyness 1936 book The General Theory of
Employment, Interest and Money was revolutionary. Unfortunately there is
not a consensus as to what was revolutionary about this volume.
230
Some economists have argued that the aggregate accounting scheme

used by Keynes was the revolutionary aspect. But Kuznets had developed
a system of aggregate income accounts by 1929, long before Keynes
was even thinking about a general theory of employment, interest and
money.
Many Old and New Keynesians believe that the Keynesian revolution is
nested in supply-side market imperfections that result in the rigidity of
money wages and prices, asymmetric information and lack of transparency. But throughout the nineteenth century classical economists had
argued that monopoly elements in the market were the cause of unemployment and Keynes (1936, chapter 19) specifically denies that such supplyside elements are the fundamental cause of unemployment. Keynes
claimed that it was on the demand side and not on the supply side that his
revolution was embedded. Surely, then, price inflexibility was not a revolutionary idea in 1936.
Many scholars, for example, Don Patinkin and Axel Leijonhufvud, have
argued that Keyness revolution was centred on the multiplier concept. Post
Keynesians believe that there is a much more fundamental foundation for
Keyness revolution. After all, if the revolutionary essence was the multiplier, then the proper name would have been the Kahnian revolution, for
Keynes merely transformed Richard Kahns employment multiplier measured in terms of employment units into an expenditure multiplier measured
in either nominal or money-wage unit terms (Keynes 1936, p. 115). It would
be hard to justify the canonization of Keynes in the economic literature if
all he had done was to focus attention on a concept that a former student
had developed and published years earlier.
Keynes (1936, p. 192) was convinced that the assumption of less than
perfect price flexibility made by the weaker [classical] spirits was not necessary to explain persistent unemployment and that this assumption caused
injury to . . . logical consistency. Instead, Keynes developed an expanded
demand classification system to demonstrate that Says Law is not the true
law relating the aggregate demand and supply functions . . . [and hence]
there is a vitally important chapter of economic theory which remains to
be written and without which all discussions concerning the volume of aggregate employment are futile (ibid., p. 26; italics added).
Says Law specifies that all expenditure (aggregate demand) on the products of industry is always exactly equal to the total costs of aggregate production (aggregate supply including gross profits). Letting Dw symbolize
aggregate demand and Zw aggregate supply (both measured in wage units,
that is, nominal values deflated by the money-wage rate), then:
Dw fd(N)
(1)
231
and
Zw fz(N).
(2)
fd(N)fz(N)
(3)
Says Law asserts that:
for all values of N, i.e., for all values of output and employment (ibid., pp.
256). In other words, in an economy subject to Says Law, the total costs
(including profits and rents) of the aggregate production of firms (whether
in perfect competition or not) are recouped by the sale of output. There is
never a lack of eective demand. The aggregate demand and aggregate
supply curves coincide (see Figure 11). In a Says Law economy, there is
never an obstacle to full employment, no matter what the degree of price
flexibility in the system.
Expected
sales,
planned
spending
Z w = Dw
Employment
0
Figure 11
A Says Law economy
To develop the true law relating Dw and Zw for a monetary economy,

Keynes produced a model where the aggregate demand and aggregate
supply functions, fd(N) and fz(N), need not be coincident (see Figure 12); as
the general case, there is no necessity for the determinants of the aggregate
demand function to be identical with the determinants of aggregate supply.
232
Expected
sales,
planned
spending
Zw
Dw
Employment
0
Figure 12
A Keynesian economy
Keynes dierentiated his theory from classical economics by a taxonomic analysis of aggregate demand. As equation (1) suggests, classical
theory fitted all expenditures into a single category, Dw, aggregate demand
(which is created entirely by supply). Keynes, on the other hand, divided all
types of expenditures into two demand classes, that is,
Dw Dw1 Dw2 fd(N)
(4)
where Dw1 represented all expenditures which depend on the level of

[current] aggregate income and, therefore, on the level of employment N
(Keynes 1936, p. 28). Thus:
Dw1 f1(N).
(5)
Logically, therefore, Dw2 represents all expenditures not related to current

income and employment:
Dw2 f(N).
(6)
Classical theory is a special case of Keyness general analytical system

that can occur only if additional axioms are imposed to force the aggregate
demand function to consist solely of expenditures equal to current income
233
at all levels of N. Demand will then have the same determinants as supply.
The necessary additional classical postulates for Says Law are:
1.
2.
3.
the axiom of ergodicity which asserts that the future can be reliably calculated from past and present market data. In Old Classical theory
ergodicity was usually subsumed when it was assumed that decision
makers possessed foreknowledge of the future. In New Classical theory
it is presumed that agents have rational expectations about a statistically reliably predictable future;
the axiom of gross substitution, so that flexible relative prices ensure
that all markets clear; and
the neutral money axiom, which ensures that changes in the nominal
money supply have no real eects.
Unfortunately, while Keynes was developing his principle of eective

demand the modern axiomatic theory of value had not yet been developed,
so that Keynes could not explicitly label the equivalents of the axiom of
parallels that had to be overthrown (Keynes 1936, p. 16) to produce a
general theory. Nevertheless, in 1937, Keynes (1937 [1973], pp. 4089) specifically noted that in the new monetary theory of production that he was
developing, the neutral money axiom was not applicable in either the short
run or the long run. Yet even today, Blanchard (1990, p. 828) proclaims that
all macroeconomic New Classical and New Keynesian models impose the
long-run neutrality of money as a maintained assumption. This is very
much a matter of faith, based on theoretical considerations [that is, axiom
based], rather than on empirical science.
Keyness specification of the essential properties of money in his
general theory requires rejecting the classical postulate that money (and all
other liquid assets) are gross substitutes for the products of industry. Money
(and all other liquid) assets possess two essential properties (Keynes 1936,
pp. 23031). These are:
1.
2.
The elasticity of production of money is zero; in essence, money is nonproducible by the use of labour in the private sector. Money does not
grow on trees. Money (and all liquid assets) therefore cannot be harvested by hiring otherwise unemployed workers to harvest money trees
whenever people demand to hold additional liquid assets as a store of
value.
The elasticity of substitution between money (that is, liquid assets) and
producible goods is zero. Accordingly, any increase in demand for
liquidity (nonproducibles to be held as a store of value), and resulting
changes in relative prices between nonproducible liquid assets and the
234

products of industry, will not divert the demand for liquidity into a
demand for goods and services. Keynes (ibid., p. 241) insisted that the
attribute of liquidity is by no means independent of these two [elasticity] characteristics and therefore as long as savers store their wealth
in assets whose elasticities of production and substitution may be very
low, unemployment equilibrium can exist no matter what the supplyside conditions are.
Since classical theory assumes that only producibles provide utility, then,
in the long run, only a lunatic would want to hold a nonproducible good
as a liquid store of value. Keynes (1936, chapter 12; 1937 [1973],
pp. 11215), on the other hand, used the concept of uncertainty to explain
why, even in the long run, people would reveal a preference to hold nonproducibles such as money as a store of value no matter how high its relative
price rose vis--vis the products of industry. (The future is uncertain rather
than merely risky in the probabilistic sense.) If nonproducibility is an essential attribute of all assets that possess the characteristic of liquidity and the
holding of liquid assets can provide a long-run security blanket against
uncertainty, then liquid assets can provide utility in a way that producibles
cannot.
Hahn demonstrated that unemployment occurs when there are in this
economy resting places for savings other than reproducible assets (1977,
p. 31) and the existence of any non-reproducible asset allows for a choice
between employment-inducing and non-employment inducing demand
(ibid., p. 39). In an uncertain world, he who hesitates to spend on producibles and holds liquid assets instead is free to make a decision another day.
By jettisoning the classical axioms of ergodicity and gross substitution,
Keynes could demonstrate that, as a general case, unemployment is possible and money is not neutral.
The axiomatic microfoundations of classical economic theory, on the
other hand, ensure that all income is always spent on the products of industry. In the simplest case all current expenditures are equal to current
income, as utility-maximizers are constrained by their income (budget-line
constraint) in their choice between good A and all other producibles. To
spend less than ones income is to reveal a preference below the budget line
and thereby to engage in non-utility-maximizing behaviour. The aggregate
of all this microfoundational spending would be classified under Dw1. The
marginal propensity to spend out of current income is unity, and any additional supply (the micro-equivalent is an upward shift in budget constraint
lines) creates its own additional demand. (In an intertemporal setting with
gross substitutability over time, agents plan to spend lifetime income on the
products of industry over their life cycle. The long-run marginal propensity
235
to spend is unity.) Consequently, in either the short run or the long run,
fd(N)fz(N) for all values of N and Figure 11 is relevant.
Keyness taxonomy was a general analysis that could lead to nonclassical results. Keyness second expenditure category, Dw2, was not equal
to planned savings (which can be defined as fz(N)f1(N)). Only if Dw2 is
assumed to be equal to planned savings is:
Dw2 fz(N)f1(N)
(7)
Dw Dw1 Dw2 f1(N)fz(N)f1(N)fz(N).
(8)
and
A comparison of equation (8) and equation (2) shows that if Dw2 is assumed
equal to planned savings, then aggregate demand and supply are identical
and Says Law holds.
To ensure that equations (7) and (8) did not represent a general case,
Keynes asserted that the future is uncertain in the sense that it cannot be
either foreknown or statistically predicted by analysing past and current
market price signals. If the future is uncertain, then expected future profits,
the basis for current Dw2 investment spending, can neither be reliably forecasted from existing market information, nor endogenously determined
from todays planned savings function, fz(N)f1(N) (Keynes 1936,
p. 210). Instead investment expenditures depend on the exogenous (and
therefore by definition, sensible but not rational) expectations of entrepreneurs, or what Keynes called animal spirits. Thus:
Dw2 f(N)
(9)
in either the short or long run.

Explicit recognition of the possibility of more classes of current demand
for producible goods and services based on a smaller axiomatic foundation
makes Keyness analysis a more general theory than classical theory. The
latter becomes a special case and not . . . the general case (ibid., p. 3), where
the category of all expenditures not related to current employment is
empty. In terms of equation (4), classical theory asserts that:
Dw2 0
(10)
Dw Dw1 f1(N)fz(N)Z
(11)
and therefore
for all values of N.
236
The next logical task for Keynes was to demonstrate that the characteristics of the special case assumed by classical theory happen not to be those
of the economic society in which we actually live (ibid., p. 3). In other
words, Keynes had to demonstrate that even if Dw2 0, the Dw1 function
would not be coincident with his macro-analogue of the age-old supply
function for all values of N. To do this Keynes had to throw over the classical axioms of neutral money (that is, the possession of money per se provides no utility) and gross substitution.
If these restrictive axioms are jettisoned, then some portion of a utilitymaximizing agents income might be withheld from the purchase of producible goods and diverted into purchasing nonproducible money and/or
other nonproducible liquid assets. The marginal propensity to spend out of
current income on the products of industry would then be less than unity.
In an uncertain world, the possession of money and other nonproducible
liquid assets provides utility by protecting the holder from fear of being
unable to meet future liabilities. As long as producible goods are not gross
substitutes for holding nonproducible liquid assets (including money) for
liquidity purposes, then no change in relative prices can induce income
earners to buy producibles with that portion of income they wish to use to
purchase additional long-run security (against non-ergodic economic conditions) by holding liquid assets.
In sum, Keyness general theory of employment must be applicable to an
uncertain (non-ergodic) world. When money and all other liquid assets
possess certain essential properties, then agents can obtain utility (by being
free of fear of possible future insolvency or even bankruptcy) only by
holding a portion of their income in the form of nonproducible liquid
assets. If the gross substitutability between nonproducible liquid assets
(including money) and producible goods is approximately zero (Keynes
1936, chapter 17; Davidson 198283, 2002), then when agents want to save
(in the form of nonproducible liquid assets) money is not neutral, even with
perfectly flexible prices. Thus, the general case underlying the principle of
eective demand is:
Dw1 f1(N)fz(N)
(12)
while planned savings, fz(N)fd(N), are equal to the amount out of current
income that utility-maximizing agents use to increase their holdings of
nonproducible liquid assets. The decision to save today means a decision
not to have dinner today. But it does not necessitate a decision to have
dinner or to buy a pair of boots a week hence or a year hence or to consume
any specified thing at any specified date (Keynes 1936, p. 210).
By proclaiming a fundamental psychological law associated with the
Keyness Treatise on Money
237
detailed facts of experience, where the marginal propensity to consume is

always less than unity, Keynes (ibid., p. 96) finessed the possibility that
equation (9) is ever applicable. If the marginal propensity to consume is
always less than unity, then f1(N) would never coincide with fz(N), even if
Dw2 0, and the special classical case is not applicable to the economic
society in which we actually live (ibid., p. 3).
In sum, Keyness principle of eective demand demonstrates that, in a
non-ergodic world, it is the existence of nonproducible assets that are held
for liquidity purposes and for which the products of industry are not gross
substitutes that is the fundamental cause of involuntary unemployment.
The lack of perfect price flexibility is neither a necessary nor a sucient
condition for demonstrating the existence of unemployment equilibrium.
P D
See also:
Bastard Keynesianism; Eective Demand; Liquidity Preference; Microfoundations; Money;
Non-ergodicity; Uncertainty; Unemployment.
References
Blanchard, O.J. (1990), Why does money aect output?, in B.M. Friedman and F.H. Hahn
(eds), Handbook of Monetary Economics, Vol. 2, New York: North-Holland, pp. 779835.
Davidson, P. (198283), Rational expectations: a fallacious foundation for studying crucial
decision making, Journal of Post Keynesian Economics, 5, Winter, pp. 18296. Reprinted in
P. Davidson, Inflation, Open Economies, and Resources, edited by Louise Davidson,
Macmillan: London, 1991, pp. 12338.
Davidson P. (2002), Financial Markets, Money and the Real World, Cheltenham, UK and
Hahn, F.H. (1977), Keynesian economics and general equilibrium theory, in G.C. Harcourt
(ed.), Microfoundations of Macroeconomics, London: Macmillan, pp. 2540.
Keynes, J.M. (1936), The General Theory of Employment, Interest and Money, New York:
Harcourt Brace.
Keynes, J.M. (1937), The general theory of employment, Quarterly Journal of Economics, 51
(2), 20923. Reprinted in The Collected Writings of John Maynard Keynes, Vol. XIII, edited
by D. Moggridge, London: Macmillan for the Royal Economic Society, 1973, pp. 10923.
All references are to the reprint.

A Treatise on Money by John Maynard Keynes was published on 24
October 1930, after six years of deep thinking and practical work on, for
example, the economic policies of the Liberal Party and the decline of the
cotton industry. In the mind of Keynes, at that time already a well-known
figure in world politics, the Treatise was designed to provide the most comprehensive and systematic analysis of monetary matters ever produced
(Harrod 1951, chapter 10). It was to do for his academic reputation what
238
political and cultural activities had already done in the public arena
(Patinkin 1976). But the Treatise has had bad luck in a way that explains
the fortunes and misfortunes of Keynesian economics more generally.
To modern students and most scholars, Keynes means The General
Theory of Employment, Interest and Money (1936). Keynes is the analyst of
unemployment and the depression. He is the theorist of aggregate demand
and equilibrium unemployment as well as the promoter of public expenditure. The Treatise is then at the best a prelude, in the words of Schumpeter,
a collection of imperfect and embarrassed first statements of General
Theory propositions (Schumpeter 1952, p. 278). According to this orthodox interpretation of Keyness work what is new and important in the
Treatise is absorbed and developed in the General Theory, and the permanent value of the latter is the explanation of short-run unemployment. An
alternative view starts with a more balanced reading of the Treatise. It
defends the originality of that work but it also emphasizes the continuous
evolution of what Keynes was later to call a monetary theory of production, that is, the search for a sound analytical framework based on the principle of non-neutrality of money (and non-neutrality of choice). This
alternative view is also supported by early interpretations of the Treatise:
The latter [the General Theory] was written somewhat in haste after Keynes had
achieved in his own mind a wide theoretical synthesis . . . he was anxious to get
this before the public quickly. The Treatise, by contrast, contains all his gathered
wisdom about monetary matters . . . It is, I would submit, impossible to have an
understanding of Keynes in depth, if one has not read the Treatise. (Harrod
1969, p. 163)
The Treatise consists of seven Books organized in two volumes, namely The
Pure Theory of Money (vol. 1, Books IIV) and The Applied Theory of
Money (vol. 2, Books VVII). Book I is concerned with the definition of
the nature of money and the description of its historical origins. Book II
deals with the laborious and complex description of various index
numbers. An important outcome of those two books is the idea that aggregate concepts like the price level or the quantity of money are not adequate
for explaining the behaviour of the main economic variables (Skidelsky
1992, chapter 10). Next follows the core of the Treatise, Books IIIIV.
There Keynes oers a formal and rigorous discussion of the static and
dynamic problems of monetary theory, including the presentation of his
fundamental equations. Book V considers the empirical magnitudes of
his key variables, and Book VI deals with the institutional features of the
banking system. Finally, Book VII lays out the implications of the analysis for both national and international stabilization policies.
The basic problem that Keynes sets out to analyse in the Treatise is the
239
instability of market economies. The analysis is very clear: the initial and
most influential cause of output fluctuations in modern economies is the
change in the level of investment. Since profit is the mainspring of those
changes, a theory of profit is then essential for the explanation of economic
fluctuations. Thus capital accumulation and income distribution are the
main themes of the Treatise (Graziani 1981). At the same time Keynes
argues that changes in the level of prices are the primary mechanism of reconciliation between the investment decisions of entrepreneurs and the distribution of income between wages and profit. These price movements are
explained by means of the fundamental equations.
The analytical framework of the Treatise is grounded on the distinction
between wage-earners and entrepreneurs, together with the related separation between expenditure on consumption goods and on capital goods:
Saving is the act of the individual consumer and consists in the negative act of
refraining from spending the whole of his current income on consumption.
Investment, on the other hand, is the act of the entrepreneur whose function it
is to make the decisions which determine the amount of the non-available
output, and consists in the positive act of starting or maintaining some process
of production or withholding liquid goods. . . . The vital point to appreciate is
this . . . the performance of the act of saving is in itself no guarantee that the
stock of capital goods will be correspondingly increased. (Book III, pp. 1558)
Saving and investment are made by two dierent groups of people, and for
Keynes there is no spontaneous market mechanism that necessarily reconciles them. It is indeed in this divergence and the consequent disequilibrium
process that Keynes envisages the source of profit creation. The story
unfolds as follows. If investment runs ahead of (below) saving it means that
entrepreneurs have decided to produce less (more) consumption goods than
the amount wage-earners have decided to purchase. The price of those
goods increases (decreases), as does the remuneration of entrepreneurs.
Abnormal (or subnormal) profit as Keynes called it is then due to the
divergence between investment and saving (equation (viii), Book III, p. 124).
The price of consumption goods plays a key role in Keyness theory of
income distribution. In equilibrium it is equal to average (long-period
normal) cost but when, as in the case above, investment diers from saving,
the price level, as determined by the first term [that is, cost], is upset by the
fact that the division of the output between investment and goods for consumption is not necessarily the same as the division of the income between
savings and expenditure on consumption (Book III, p. 123). Changes in the
price level of consumption goods have the role of reconciling, via the mechanism of forced saving (involuntary abstention), the production decisions
of entrepreneurs with the expenditure plans of wage-earners. For instance,
240
with investment running ahead of saving and under the assumption that
abnormal profit is entirely saved, a redistribution of income from entrepreneurs to wage-earners takes place and equilibrium is again restored
(Kahn 1984, pp. 678). Keynes formally derived these results from his definition of the price level of consumption goods as equal to the monetary
cost of production per unit of output (what Keynes called the rate of eciency earnings) plus any element of abnormal (or subnormal) profit per
unit of output. This is the meaning of the first fundamental equation; the
second fundamental equation extends this idea to the price level of output
as a whole (Book III, pp. 1224).
Thus booms and slumps derive from the dierence between saving and
investment. More importantly for what would later be argued in the
General Theory, Keynes argues that voluntary abstinence is no guarantee
of prosperity. This idea is well illustrated in the famous banana parable
describing a community in a state of equilibrium that is disturbed by a
thrift campaign (Book III, chapter 12, sec. ii). With an increase in saving
entrepreneurs are now forced to reduce the price of bananas (otherwise
they will rot). The unexpected (abnormal) loss causes entrepreneurs to
reduce their wage bills by laying o workers and/or cutting wages. But this
does not help as long as the community continues to save more than it
invests. Output declines until either the thrift campaign is called o or
peters out as a result of the growing poverty; or . . . investment is stimulated
by some means or another so that its cost no longer lags behind the rate of
saving (p. 160). A similar case, of course, could be argued for an increase
of investment in, for example, new banana plantations. The crux of the
whole parable is that saving can be made identical to investment as a result
of a change in output though, as Robinson argued, Keynes may have failed
fully to realize that he had actually elaborated a long-period analysis of
output (Robinson 1960, p. 56).
In terms of the policy implications of the analysis, Keynes suggests that
the banking system should try to eliminate economic fluctuations and in
consequence to stabilize the purchasing power of money. By keeping the
market rate of interest equal to the natural rate of interest, that is, to the
notional rate at which saving and the value of investment are exactly balanced (Book III, p. 139), the banking system would preserve the condition
of zero (abnormal) profit. Keynes believed that in general the banking
system could do the job. However, he also warned that the natural rate of
interest relies on profit expectations of entrepreneurs, and hence it is independent of the objective technical conditions of production. A potential
conflict may then arise between domestic and external equilibrium. In a
severe slump, with profit expectations running very low, the banking system
may not be able to aord eective reductions in the market rate without
241
causing a large outflow of domestic currency. In this case Keynes suggested

that the Government must itself promote a programme of domestic investment (Book VII, p. 337).
Several authors have argued that the Treatise and the General Theory
embody the same vision, the same appreciation of empirical observation
and the same analytical structure (for example, Amadeo 1989, chapters
34). What really dierentiates the Treatise from the General Theory is the
formal method and the specific purpose of the analysis. Keynes had always
in mind a close connection between theory and practice, but under the pressure of historical events his thought about the form of that connection went
on changing. In 1930 Keynes was mainly concerned with the instability of
market economies, the ups and downs that characterize the credit cycle.
Output and employment were seen as moving around some norm, and he
aimed to explain the causal mechanism behind those movements. But by
1936 he felt that the problem was now not with fluctuations around a norm
but with the norm itself. Persisting mass unemployment was the practical
interest, and for the sake of getting his solution across, he was content to
set aside some of his most brilliant though highly heterodox ideas. This is
the ultimate link between his two major books and the main reason for
reverting to the Treatise.
G F
See also:
Business Cycles; Growth and Income Distribution; Monetary Policy; Money; Saving.
Bibliography
Amadeo, E.J. (1989), Keyness Principle of Eective Demand, Aldershot, UK: Edward Elgar.
Graziani, A. (1981), Keynes e il Trattato sulla Moneta [Keynes and the Treatise on Money],
in A. Graziani, C. Imbriani and B. Jossa (eds), Studi di Economia Keynesiana [Studies in
Keynesian Economics], Naples: Liguori, pp. 21134.
Harrod, R.F. (1951), The Life of John Maynard Keynes, London: Macmillan.
Harrod, R.F. (1969), Money, London: Macmillan.
Kahn, R.F. (1984), The Making of Keyness General Theory, Raaele Mattioli Lectures,
Moggridge, D.E. (1992), Maynard Keynes: An Economists Biography, London: Routledge.
Patinkin, D. (1976), Keyness Monetary Thought. A Study of Its Development, Durham, NC:
Duke University Press.
Robinson, J. (1933 [1960]), The theory of money and the analysis of output, in J. Robinson,
Collected Economic Papers, Vol. I, Oxford: Basil Blackwell, pp. 528.
Schumpeter, J.A. (1946 [1952]), John Maynard. Keynes, in J.A. Schumpeter, Ten Great
Economists, London: Allen & Unwin, pp. 26091.
Skidelsky, R. (1992), John Maynard Keynes: The Economist as Saviour 19201937, London,
Macmillan.
Liquidity Preference
The idea that the rate of interest adjusts until the supply of savings is
brought into equality with the demand for savings dates back at least as far
as Henry Thornton. The proposition is exemplified by the loanable funds
theory of the rate of interest, which Keynes attempted to overthrow in The
General Theory. Ralph Hawtrey, Bertil Ohlin and Dennis Robertson were
Keyness primary adversaries, each subscribing to the idea that the rate of
interest would settle at the point where the (flow) supply of savings and the
(flow) demand for investible funds would be equal. According to the theory,
an increase in savings (given the level of income) would bring the rate of
interest down just enough to stimulate capital production to the point
where the additional saving would be exactly exhausted by the additional
demand for new investment. Accordingly, an increase in aggregate saving
could alter the composition of aggregate demand (that is, less consumption
and more investment), but it could not cause a reduction in the level of
aggregate demand. Thus, full employment could not be undermined by the
preference of households to save rather than spend money.
The experience of the Great Depression in the 1930s made it impossible
to sustain the notion that capitalist economies were self-regulating. Moreover, it caused many economists to become dissatisfied with the quantity
theory of money the theory that holds that changes in the quantity of
money are primarily responsible for causing changes in national income. By
the logic of the quantity theory, it should have been possible for the central
bank to stop the downward spiral by increasing the money supply. Keyness
liquidity preference theory oered a new perspective on these problems.
The dierence between Keyness analysis and the traditional theory of
interest, which he considered a nonsense theory (Keynes 1936 [1964],
p. 177) essentially amounts to a distinction between the economics of full
employment and the economics of unemployment. In Keyness view, the
traditional analysis was flawed because it treated saving and investment as
the determinants of the system and the rate of interest as a determinate
when, in fact, savings and investment are the determinates of the system,
and the rate of interest is a determinant. Moreover, he emphasized that
because saving depends upon income and income depends upon investment, it was impossible to conceive of an independent shift in either schedule. According to Keynes, a shift in the savings schedule would, in general,
cause a change in income, with the result that the whole schematism based
on the assumption of a given income breaks down (ibid., p. 179).
242
Liquidity preference 243

An equally crucial blow to the traditional theory followed from Keyness
recognition that money, held in its barren form, yields no monetary reward.
According to the loanable funds theory, interest was the reward for parting
(today) with command over goods and services. But, as Keynes pointed out,
the act of saving, by itself, generates no necessary reward in the form of
interest. Wealth can be stored in various forms (for example, idle balances,
short- or long-term financial instruments, capital assets, or other forms of
income-earning property), but interest will be forthcoming only if individuals agree to part with idle cash in favour of short- or long-term financial
instruments. With this in mind, Keynes set out to discover why anyone,
outside of a lunatic asylum, would ever choose to hold money in its barren
form.
Initially, he oered three reasons: (i) the transactions motive, (ii) the precautionary motive and (iii) the speculative motive. The transactions motive
refers to a desire to hold money for the purpose of meeting anticipated
expenditures (that is, bridging the time gap between known receipts and
known expenditures). The precautionary motive, in contrast, has to do with
the desire to hold cash for the purpose of meeting unanticipated expenditures. According to Keynes, increases (decreases) in the level of income
would be associated with increases (decreases) in the size of the cash balances held to satisfy the transactions and precautionary motives. Like the
precautionary motive, the speculative motive arises as a consequence of
uncertainty. However, unlike the transactions and precautionary motives,
an individuals desire to hold money to satisfy the speculative motive is a
function of anticipated movements in a range of asset prices rather than
changes in the level of income. Here, the idea is that individuals hold money
in order to hedge against a declining securities market. If, for example, bond
prices are expected to rise, individuals will prefer to become less liquid today
(that is, to transform idle money balances into bonds) in order to take
advantage of anticipated capital gains. In contrast, if speculators expect
bond prices to fall to a point where interest gains are outweighed by capital
losses, they will prefer to become more liquid today (that is, to hold idle
money balances instead of less liquid financial instruments). The money
balances that are held to satisfy these three motives reflect the individuals
degree of liquidity preference.
The desire to hold liquid assets (that is, coins, paper currency and bank
deposits) can be analysed graphically by means of a liquidity preference
function. Leaving aside the desire to hold money to satisfy the transactions
and precautionary motives, Keynes proposed that the liquidity preference
function could be envisaged as a continuous curve relating changes in the
demand for money to satisfy the speculative motive and changes in the rate
of interest as given by changes in the prices of bonds and debts of various
244
Liquidity preference
maturities (ibid., p. 197). Figure 13 depicts this relation. The liquidity function La reflects the desire to hold money as an asset (that is, the desire to
hold idle cash balances to satisfy the speculative motive). The curve shows
the quantities of money that individuals, in the aggregate, will wish to hold
at various rates of interest.
i
La = f(i)
0
Figure 13
La
The liquidity preference function
Keynes gave two reasons for the downward-sloping nature of the liquidity function. First, he explained that a fall (rise) in the rate of interest will
be associated with an increase (decrease) in the demand for money as an
asset, since the penalty for being liquid (that is, the forgone interest) has
been reduced. Second, he suggested that a decline (rise) in the rate of interest will be associated with an increased (decreased) demand for cash balances to satisfy the speculative motive because some market participants
will anticipate a future rise (fall) in the rate of interest.
As Keynes explained, both the position and the shape of the liquidity
preference function depend upon the state of long-term expectations. What
matters most are expectations regarding the future path of asset prices.
Thus if speculators, who are constantly forming expectations about the
future course of interest rates, believe that the markets estimate of future
interest rates (as implied by current prices) is mistaken, they will attempt to
profit from knowing better what the future will bring. If they believe that
future rates will lie below those currently assumed by the market, they will

have an incentive to reduce their liquidity by borrowing money today in
order to purchase longer-term bonds. In contrast, if they expect future rates
to lie above those assumed by the market, they will have an incentive to
increase the liquidity of their portfolios by selling bonds today.
Only when the expected return on every financial asset in every portfolio
is equal will market participants be satisfied with their existing holdings.
Thus, in order for the money market to be in equilibrium, expectations must
be such that investors are willing to hold the current mix of money and
bonds, given the current structure of interest rates. In contrast to the loanable funds theory of the rate of interest, which treats the rate of interest as
the outcome of the forces equilibrating the flow supply of loanable funds
with the flow demand for investible resources, this complex of rates will be
determined by the stock demand for money relative to its stock supply.
Given the aggregate money supply, only the portion not desired to satisfy
the transactions and precautionary motives is available to satisfy the speculative motive (that is, the stock demand for money). Figure 14 incorporates this constraint into the analysis. Ma is drawn as a vertical schedule,
indicating the quantity of money available to satisfy the speculative motive.
The interaction of these schedules determines not just the rate of interest
on money but also the complex of rates on assets of various maturities.
These rates can change either because the supply of money available to
satisfy the speculative motive has been altered (that is, Ma shifts) or because
i
Ma
i*
La = f(i)
0
Figure 14
La, Ma
Determination of the rate of interest
246
the expectations that determine the position and shape of the liquidity preference function have changed (that is, La shifts).
At any interest rate above i*, the stock of money available to be held as
an asset exceeds the amount of money that individuals, in the aggregate,
wish to hold to satisfy the speculative motive. In this situation, the rate of
interest will have to fall until all of the money that is not desired for transactions or precautionary balances is willingly held to satisfy the speculative
motive. Similarly, at any rate of interest below i*, the quantity of money
that market participants wish to hold as an asset exceeds the amount of
money available to be held as an asset. Assuming the banking system and
the monetary authority do not respond by increasing the money supply, the
rate of interest will have to rise to the point where the public is willing to
hold the existing stock of money (or, more appropriately, the stock of existing securities). Thus the rate of interest is the premium required to induce
market participants to hold less-liquid assets.
Two important conclusions distinguish Keyness analysis from the
(neo)classical approach, which incorporated both loanable funds theory
and the quantity theory of money. First, there is no reason to suppose that
the rate of interest will settle at a price that is consistent with full employment. According to Keynes, saving and investment determine not the rate
of interest but the aggregate level of employment. Moreover, an increase in
the propensity to save (for example, a rise in liquidity preference) should be
expected to increase interest rates, discourage investment and reduce
employment not reduce the interest rate and induce enough investment to
maintain full employment. Second, changes in the money supply have a
direct eect on interest rates (through portfolio adjustments) and only an
indirect eect on national income. This does not imply that the monetary
authority can simply manipulate the money supply until an interest rate
consistent with full employment is achieved; changes in the state of longterm expectations can always override any desired policy outcome. Keynes
makes this clear in his attack on the quantity theory argument, cautioning
that if . . . we are tempted to assert that money is the drink which stimulates the system to activity, we must remind ourselves that there may be
several slips between the cup and the lip (ibid., p. 173).
Generally speaking, Keyness theory of liquidity preference remains an
integral part of the Post Keynesian approach. But there are some notable
exceptions. Basil Moore (1988), for example, argues that the liquidity preference theory actually undermines the rationale of The General Theory.
According to Moore, liquidity preference theory is incompatible with
endogenous monetary theory. Lavoie (1985) agrees, arguing that Post
Keynesians must reject liquidity preference theory because it is relevant
only in the context of an exogenously determined money supply.

The criticisms of Moore and Lavoie appear to follow from the manner
in which Keynes treated the determination of the money supply in The
General Theory. According to Moore, Keynes should have retained the
endogenous treatment of the money supply he adopted in his Treatise on
Money and designated the interest rate as an exogenously determined variable. If the interest rate is exogenously determined (by the monetary
authority) and the supply of money is perfectly elastic at this rate, then an
outward shift of the liquidity preference function cannot increase the rate
of interest. Thus, for Moore and Lavoie, the problem appears to be this: if
the supply of money always fully accommodates the demand for money
(that is, the money supply is endogenous), then it cannot be true that an
increase in the demand for money will increase the rate of interest.
Wray (1990) argues that liquidity preference and endogenous money are
indeed compatible and that the criticisms of Moore and Lavoie are based
on a failure to distinguish between the (stock) demand for money and the
(flow) demand for credit. He, Kregel (1986), Davidson (1978) and Dow and
Dow (1989) emphasize the finance motive, which Keynes introduced in
1937 as a fourth motive for holding money. In this 1937 article, Keynes reasserts his commitment to the endogenous theory of money, explaining that
as long as banks are willing to accommodate an increase in the demand for
credit, there is no reason why the interest rate should rise with an increase
in the level of planned activity. Thus, while changes in the (stock) demand
for money can aect the rate of interest, changes in the (flow) demand for
credit (that is, bank borrowing) need not.
When the concept of liquidity preference is extended to banks and other
financial institutions, the rate of interest will be determined by the supply
of money (as determined by the banking system) and the liquidity preference of both commercial banks and the non-bank public. Wray (1990)
makes an excellent case for the compatibility of liquidity preference and
endogenous money and provides a thorough review of the role of liquidity
preference in the general theories of both Keynes and the Post Keynesians.
S B
See also:
Banking; Endogenous Money; Expectations; Money; Uncertainty.
References
Dow, Alexander C. and Sheila C. Dow (1989), Endogenous money creation and idle balances, in John Pheby (ed.), New Directions in Post-Keynesian Economics, Aldershot:
Edward Elgar, pp. 14764.
248
Keynes, J.M. (1937), Alternative theories of the rate of interest, Economic Journal, 47 (186),
24152.
Kregel, Jan (1986), A note on finance, liquidity, saving, and investment, Journal of Post
Lavoie, Marc (1985), Credit and money: the dynamic circuit, overdraft economics, and Post
Keynesian economics, in Marc Jarsulic (ed.), Money and Macro Policy, Boston, Dordrecht
and Lancaster: Kluwer-Nijho Publishing, pp. 6384.
Moore, Basil (1988), Horizontalists and Verticalists: The Macroeconomics of Credit Money,
Wray, L. Randall (1990), Money and Credit in Capitalist Economies: The Endogenous Money
Approach, Aldershot: Edward Elgar.
Marginalism
Marginalism is a term used to refer to the process of developing theoretical propositions from the imposition of conditions on the marginal values
of variables. In modern neoclassical economics these conditions are derived
from the assumption that economic agents engage in optimizing behaviour.
In particular, the conditions are expressed as first-order conditions for optimization of the agents objective function, along with the corresponding
second-order conditions to ensure a maximum or minimum as appropriate
to the problem. A standard example is derivation of the proposition that
product supply curves are upward sloping in perfectly competitive markets.
This proposition follows from finding that product price equal to marginal
cost is the solution to the first-order condition for profit maximization by
individual producers, while rising marginal cost with output is required to
satisfy the corresponding second-order condition assuming a horizontal
demand curve facing the producer.
The use of marginal conditions to develop theoretical propositions has
a long tradition in economics. An early example is David Ricardos analysis of the distribution of income between workers, capitalists and landowners in The Principles of Political Economy and Taxation (1817). Here,
Ricardo determines the rent of land of varying quality by imposing the
marginal condition that the rent of land at the extensive (or intensive)
margin equal zero. He then imposes the conditions that capital accumulates
until profits equal zero and that the wage of workers is driven to the subsistence level by the growth of population. In modern parlance, the growth
of capital and population are each governed by a marginal condition.
Marginalism was established as an all-encompassing method of theoretical analysis in economics with the marginalist revolution of the early neoclassical economists in the late nineteenth century. In the writings of Carl
Menger, W.S. Jevons, Lon Walras and Vilfredo Pareto, Ricardos analysis
of production and supply based on conditions at the margin was extended
by an analysis of demand conditions based on marginal utility. Alfred
Marshall consolidated the revolution in his Principles of Economics (1890),
providing a short-period marginal analysis of supply and demand to complement Ricardos long-period supply analysis.
Modern neoclassical economics extends marginalism from its primary
use in theoretical analysis to encompass also empirical analysis. Machlup
(1946) argues that the lack of realism in the assumptions of marginalism
does not undermine its usefulness as a technique for predicting behaviour.
249
250
Marginalism
To Machlup, decision makers can, and do, act as if they are equating marginal conditions, even though they do not have the information necessary
to calculate exact values. Errors in optimization are to be expected, but the
way in which changes in the essential variables will aect the probable decisions and action of the business man is not much dierent if the curves
which the theorist draws to depict their conjectures are a little higher or
lower, steeper or flatter (ibid., p. 536).
Today, marginal analysis dominates mainstream economics in both
theory and practice. Students around the world are taught that marginal
analysis constitutes the principles of economics. Applications of economics
to business decisions and public policy are primarily based on evaluating
changes in marginal conditions associated with hypothesized behaviour.
Finally, estimation of economic relationships is generally based on assuming that economic agents optimize according to the solution of marginal
conditions, but that they make errors that are normally distributed with zero
mean.
Criticism of marginalism is fundamental to Post Keynesian economics.
Indeed, the Keynesian revolution began with Keyness critique of the application of marginal conditions for equilibrium in the labour market in The
General Theory of Employment, Interest and Money. Keynes accepts the
application of marginal analysis to the employment decisions of individual
businesses and the labour supply decisions of workers, but argues persuasively that the marginal condition of wage equal to the value of the marginal product of labour is not sucient to guarantee full employment in a
capitalist economy. The problem noted by Keynes is that when wages and
prices move together, a reduction in the money wage cannot be relied on to
remove excess supply of labour. What is required is an increase in eective
demand, which is most readily accomplished by government expenditure
programmes.
Keyness followers at Cambridge extended the critique of marginalism to
other notions of equilibrium based on marginal conditions. Roy Harrod
pointed to the fragility of full employment in a growing economy due to
the knife-edge requirements for a balance between the growth in supply and
growth in eective demand. Joan Robinson attacked the neoclassical
parable of accumulation, noting that it ignores changes in relative prices
required to maintain the marginal conditions for equilibrium in input
markets with a change in the ratio of interest rate to wage rate. According
to Robinson, these changes in relative prices destroy our ability meaningfully to measure a quantity of capital from the collection of heterogeneous
productive equipment and structures used in the production process.
The attack on marginalism by Keynes and his followers emphasizes
problems with its internal coherence, particularly problems encountered
Marginalism
251
once marginal conditions for individual optimization decisions are aggregated to examine the behaviour of an economic system in either a static
context or through time. Other Post Keynesians go further and reject the
usefulness of marginalism in its entirety. A seminal contribution to this
strand of critique is the pioneering study of business pricing behaviour by
Hall and Hitch (1939). Hall and Hitch surveyed business executives responsible for price setting and concluded that prices are predominantly set by
reference to average cost, suggesting the need for a pricing theory based on
the full-cost principle rather than a theory based on the marginal conditions required for profit maximization. Machlups (1946) argument about
economic agents acting as if they are engaged in optimizing behaviour was
aimed directly at refuting the survey evidence provided by Hall and Hitch.
Both marginalists and followers of the full-cost principle now understand that there can be formal equivalence between a profit-maximizing
price and a full-cost price. Indeed, there is recognition that many influences,
such as increases in prices of variable inputs, have a similar impact on prices
in both approaches. However, as Lee (199091, p. 263) notes, it is clear that
the Post Keynesian full cost pricing equation is radically dierent from its
marginalist counterpart. Particularly important to Lee as distinctive features of Post Keynesian pricing are (i) the absence of an influence of
demand on prices, (ii) a role for social conventions in determining prices
and (iii) the evolving nature of the firm such that prices and quantities have
a historical/temporal dimension that eectively precludes a determinant
relationship between price and quantity, and price-quantity and maximizing objectives (ibid., p. 259).
In spite of the general hostility towards marginalism, some concepts with
origins in marginal analysis have survived and prospered in Post Keynesian
literature. A good example is the degree of monopoly. Lerner (1934) develops the degree of monopoly as the dierence between price and marginal
cost divided by price as a way of measuring the impact of monopoly on the
eciency of resource allocation, thereby providing a clear application of
marginalism in economics. Lerners definition is adopted in Kaleckis
(1938) analysis of the distribution of income. Further, Kalecki (1940) subsequently provides his own derivation based on satisfying the marginal conditions for profit maximization under conditions of imperfect competition.
However, Kriesler (1987) suggests that Kaleckis disenchantment with
aspects of the marginalist derivation eventually led him to substitute a distinctly non-marginalist pricing equation as the basis for the degree of
monopoly in his analysis of income distribution.
In summary, marginalism has not been banished completely from Post
Keynesian economics, in spite of strong criticism by some Post Keynesians
and outright rejection by others. This diversity of treatment is consistent
252
Microfoundations
with the open systems ontology espoused by Post Keynesian methodologists. Rather than seek sole allegiance to a single method of analysis, as is
the case with marginalism for our neoclassical colleagues, Post Keynesians
recognize the virtue of using a variety of analytical methods in furthering
our understanding of a complex and continually evolving economic
system. In this context, Downwards (1999, chapter 6) assessment of the
debate between marginalists and advocates of the full-cost pricing principle is particularly revealing.
H B
See also:
Capital Theory; Employment; Income Distribution; Pricing and Prices; Walrasian
Economics.
References
Downward, Paul (1999), Pricing Theory in Post Keynesian Economics, Cheltenham, UK and
Hall, R.L. and C.J. Hitch (1939), Price theory and business behaviour, Oxford Economic
Papers, 2, 1245.
Kalecki, Michal- (1938), The determinants of the distribution of national income,
Econometrica, 6 (2), 97112.
Kalecki, Michal- (1940), The supply curve of an industry under imperfect competition,
Review of Economic Studies, 7, 91112.
Kriesler, Peter (1987), Kaleckis Microanalysis, Cambridge: Cambridge University Press.
Lee, Frederic S. (199091), Marginalist controversy and Post Keynesian price theory, Journal
of Post Keynesian Economics, 13 (2), 24963.
Lerner, Abba P. (1934), The degree of monopoly, Review of Economic Studies, 1, 15775.
Machlup, Fritz (1946), Marginal analysis and empirical research, American Economic
Review, 36 (4:1), 51954.
Microfoundations
The term microfoundations has entered discussions of Post Keynesian
theory from neoclassical analysis. In its broadest sense, the term refers
simply to the behavioural specification for individual economic agents
within macroeconomic models. With this definition, all behavioural theories of individual behaviour within macroeconomic models constitute
microfoundations.
The practical meaning of the term in neoclassical analysis, however, is
more narrow. In common mainstream usage, the term microfoundations
refers to the linkage between macroeconomic models and the maximization
of utility and profit by individual agents. While this procedure dates back
at least to the life-cycle model of consumption (Franco Modigliani) and the
neoclassical model of investment (Dale Jorgenson), the new classical
Microfoundations
253
macroeconomic revolution of the 1970s placed much greater emphasis on

microfoundations in this narrow sense. The reason for this methodological
shift is the famous Lucas critique. This idea arose initially from models
that explore the neutrality of money with rational expectations, but its
implications have spread far beyond this context. The Lucas critique questions the usefulness of the predictions from any economic model that is not
specified in terms of so-called deep, structural parameters of taste and technology, for only these parameters are invariant to dierent policy regimes.
This dictum has been used in much of neoclassical macroeconomics to
dismiss a priori any analysis, theoretical or empirical, without microfoundations that explicitly link results to parameters of tastes through utility
maximization and parameters of technology through profit maximization.
The interface between this approach and results in Post Keynesian economics is very large. With respect to methodology, two issues are salient:
the role of optimization and the specification of expectations. Post
Keynesian writers have an eclectic approach to optimization. There are
numerous examples of the use of optimization as the foundation for behavioural theory in Keynesian and Post Keynesian research. In discussing the
classical postulates in chapter 2, section V of the General Theory, Keynes
accepted the view that the marginal productivity of labour equals the real
wage. Davidson (1998) describes the point of eective demand in terms of
the behaviour of profit maximizing entrepreneurs. Fazzari et al. (1998) use
maximization for a representative firm to explore the impact of imperfect
competition on firms employment and pricing decisions in a Keynesian
macroeconomic model. Keyness analysis of the marginal eciency of
capital in chapter 11 of the General Theory concludes that investment will
equate the marginal eciency of the capital stock to the market rate of
interest, consistent with the results of a maximization model. Unlike the
new classical approach, however, there is no methodological restriction in
Post Keynesian analysis that dictates the use of optimization as a behavioural axiom. Many relations are motivated by behaviour that does not rely
solely on optimization. For example, Keyness fundamental psychological
law links consumption to income without reference to any explicit optimization conditions. However, as the examples above show, the use of optimization to explore microfoundations need not be inconsistent with the Post
Keynesian research programme.
A more significant dierence between neoclassical and Post Keynesian
microfoundations arises in the specification of expectation formation. The
neoclassical rational expectations approach specifies expectations from the
actual statistical predictions of the model. Expectations therefore have no
independent existence in such models; they are determined fully by the structural microfoundations, usually preferences and technology. This modelling
254
Microfoundations
approach contrasts sharply with the great emphasis put on the significant
and independent role for expectations in Keynesian and Post Keynesian
research. In these models, changes in expectations can occur without any
change in technology or preferences, and such independent changes in
expectations significantly aect the predictions of the model. This approach
allows phenomena such as convention, confidence and animal spirits
to aect macroeconomic outcomes. Although long-standing work on
bounded rationality has received some recent attention among neoclassical macro economists, there remains a great gulf in the microfoundations of
expectations between the modern neoclassical tradition, which usually
strives to make expectations entirely subordinate to technology and preferences, and the Post Keynesian approach, in which expectation formation is
itself an independent and fundamental influence on the path of the
economy. (Crotty 1994 makes a similar point and links it insightfully to an
assessment of the logic of grounding neoclassical models in the microfoundations of individual agent choices.)
This gulf in the conception of expectation formation reflects more than
a dierence in modelling strategy. It is linked inherently to the significant
dierence between orthodox and Post Keynesian conceptions of uncertainty. Uncertainty in neoclassical macroeconomic models is represented
by probability distributions over ergodic stochastic processes. Most Post
Keynesian writers reject this conception in favour of fundamental, true
or Keynesian uncertainty, which implies that stable, objective probability
distributions do not exist for key variables that aect microeconomic decisions. The microfoundations question for expectations therefore is not
whether real-world agents do or do not behave in accordance with the rational expectations hypothesis. In the Post Keynesian perspective, there is no
logical basis for rational expectations because the statistical information
required for rational expectations simply does not exist.
Clearly, the choice of microfoundations plays a key role in understanding the dierence between Post Keynesian and other forms of economic
analysis. The remainder of this entry considers how these dierences
appear in three dierent contexts of central significance for macroeconomics: labour demand, the financing of investment and the role of nominal
rigidity.
The first of Keyness classical postulates (General Theory, chapter 2)
presents a classical theory of labour demand that appears consistent with
the modern neoclassical microfoundations approach. That is, firms maximize profits constrained by their technology. Labour demand arises from
the solution to the first-order condition for maximization that equates the
physical marginal product of labour to the real wage. Because Keynes
maintains the first classical postulate, it appears on the surface that his
Microfoundations
255
theory is consistent with at least this aspect of neoclassical microfoundations. Such a conclusion, however, is misleading. While the Keynesian
theory is consistent with the result that the real wage equals the marginal
product of labour at an eective demand equilibrium, this result diers
from the conclusion that the technologically-determined marginal product of labour is the labour demand schedule. For as Davidson (1998,
p. 825) and Fazzari et al. (1998, p. 534) argue, one cannot define the eective demand for labour in a Keynesian model without knowledge of the
state of aggregate demand and its impact on firms output markets. In a
Keynesian model, the eective demand for labour is a single point on the
marginal product of labour schedule, the point consistent with constraints on firm sales imposed by the state of aggregate demand. Fazzari
et al. generalize this result to the empirically relevant case of imperfect
competition. They argue that the presence of market power at the micro
level makes the aggregate sales constraints arising from insucient aggregate demand evident in the microfoundations of firm behaviour. Furthermore, the importance of sales constraints for microeconomic production
and employment decisions need not come from price or wage rigidities.
These results contrast strongly with the neoclassical analysis of labour
demand and production, for which aggregate demand is irrelevant, at
least in the absence of nominal rigidities. These dierences do not stem
from dierences in the behavioural postulates in the microfoundations,
such as profit maximization. Rather, the distinction arises from a dierent conception of how microeconomic agents are embedded in the macroeconomic environment.
The study of the link between finance and investment provides another
example of an active research area in which the distinction between Post
Keynesian and neoclassical microfoundations plays an important role.
Keynes and many Post Keynesian authors, perhaps most prominently
Hyman Minsky, have emphasized the connection between financial markets,
the availability of financing, and investment spending. When Jorgenson and
others developed neoclassical microfoundations for investment, however,
any consideration of financial constraints disappeared. Jorgenson explicitly
linked this approach to the ModiglianiMiller theorem that gives conditions
under which real firm investment decisions are independent of financial
structure. This research eectively eliminated financial considerations from
most mainstream investment research for two decades. In the 1980s, however,
a deeper analysis of the microfoundations of investment decisions with
asymmetric information between firms and lenders led to reconsideration of
the independence of investment from finance. The result has been a large
mainstream literature that supports the importance of financing constraints
for investment. This analysis of the microfoundations of investment with
256
Microfoundations
asymmetric information appears to bring mainstream investment theory

closer to the Post Keynesian perspective of Minsky and others.
The extent of convergence between neoclassical and Post Keynesian
microfoundations for financial eects on investment has generated controversy. There is little doubt that the large body of mainstream empirical
research that supports the importance of financial conditions for investment also supports Post Keynesian views. There is more debate on the
theoretical side, regarding the importance of asymmetric information for
the microfoundations of financing constraints. Fazzari and Variato (1994,
pp. 35961) point out that meaningful financing constraints on firm
investment require that potential lenders refuse to fund investment projects that firms wish to undertake. They conclude that this outcome
implies a systematic asymmetry in the information lenders and firms have
about investment projects. Furthermore, they argue that asymmetric information is not a minor imperfection in real-world capital markets. Rather
it is a pervasive characteristic of decentralized market economies and it
therefore must be part of the microfoundations of investment theory.
Along similar lines, Dymski (1993) discusses the complementarity of
asymmetric information and Post Keynesian microfoundations based on
a Keynesian conception of uncertainty. In contrast, however, Van Ees and
Garretsen (1993) argue that microfoundations based on asymmetric information, at least the kind of asymmetric information models that have been
pursued in mainstream analysis, cannot be reconciled with Post Keynesian
theory because the mainstream models require an ergodic stochastic environment which is inconsistent with fundamental uncertainty. Crotty
(1996) also explores the importance of expectations for investment with
fundamental uncertainty. He argues that many important aspects of
Keynesian investment theory and the linkages between investment and
finance cannot be understood from microfoundations of asymmetric
information models.
Finally, the mainstream literature known as New Keynesian is largely
motivated by the exploration of microfoundations that are of little relevance
to Post Keynesians. The New Keynesian approach emphasizes nominal
rigidity of either wages or prices. Following the elevation of optimizing
microfoundations after the Lucas critique, simply assuming nominal rigidity was considered ad hoc. New Keynesian macroeconomics therefore strives
to explain nominal rigidity with optimizing models. This class of microfoundations has little relevance for Post Keynesian macroeconomics
because the Post Keynesian approach does not require nominal rigidity to
obtain eects of aggregate demand on output and employment (see Fazzari
et al. 1998 for discussion and further references).
S F
Monetary policy 257

See also:
Agency; Credit Rationing; Eective Demand; Employment; Expectations; Financial
Instability Hypothesis; Investment; Marginalism; New Classical Economics; New Keynesian
Economics; Non-ergodicity; Uncertainty; Wages and Labour Markets.
References
Crotty, James R. (1994), Are Keynesian uncertainty and macrotheory incompatible?
Conventional decision making, institutional structures and conditional stability in
Keynesian macromodels, in G. Dymski and R. Pollin (eds), New Perspectives in Monetary
Macroeconomics: Explorations in the Tradition of Hyman Minsky, Ann Arbor: University
of Michigan Press, pp. 10542.
Crotty, James R. (1996), Is New Keynesian investment theory really Keynesian? Reflections
on Fazzari and Variato, Journal of Post Keynesian Economics, 18 (3), 33357.
Davidson, Paul (1998), Post Keynesian employment analysis and the macroeconomics of
OECD unemployment, Economic Journal, 108 (440), 81731.
Dymski, Gary (1993), Keynesian uncertainty and asymmetric information: complementary
or contradictory?, Journal of Post Keynesian Economics, 16 (1), 4954.
Fazzari, Steven M., Piero Ferri and Edward Greenberg (1998), Aggregate demand and firm
behavior: a new perspective on Keynesian microfoundations, Journal of Post Keynesian
Economics, 20 (4), 52758.
Fazzari, Steven M. and Anna Maria Variato (1994), Asymmetric information and Keynesian
theories of investment, Journal of Post Keynesian Economics, 16 (3), 35169.
Van Ees, Hans and Harry Garretsen (1993), Financial markets and the complementarity of
asymmetric information and fundamental uncertainty, Journal of Post Keynesian
Economics, 16 (1), 3748.
Monetary Policy
Whatever its detailed variants, Post Keynesian monetary policy is unified
by four characteristics, three of which stem directly from Keyness General
Theory. Of the three, the first is the dominant role given to aggregate
demand in determining the level of economic activity; the second is the lack
of any automatic tendency for that level of economic activity to coincide
with the full-employment level of output; the third is the desire for a more
equal distribution of income and wealth. The fourth characteristic, not
present in the General Theory itself, is the acceptance of an endogenous
money supply where the role of the central bank consists of setting the
price at which it will make available the reserves required to validate the
lending which banks wish to undertake, given the demand for loans. This
price is the central banks ocial dealing rate and it provides the foundation for the level and structure of market rates. In the light of these interest rates and what is often called the state of trade borrowers decide on the
(flow) of new bank lending that they require and the resulting new loans
create additional deposits, that is, money. Thus monetary policy in Post
Keynesian economics is interest rate policy (see, for example, Moore 1988).
In the Post Keynesian view, monetary policy aects both output and
258
Monetary policy
prices. Moreover, the eect upon prices runs through output. This is quite
dierent from the more extreme versions of neoclassical economics where
monetary policy is seen to aect only prices, and diers also from the more
moderate versions where output may also be aected in the short run but
as an unfortunate, simultaneous, byproduct of interest changes.
The immediate eect of a rise in interest rates falls upon demand,
through five channels: a wealth eect as a result of falling asset prices; an
increase in the price of consuming now as opposed to saving; a rise in the
cost of credit; an increase in the external value of the currency and a redistribution of spending power from the relatively poor to the better-o with
higher propensities to save. Since firms are largely price-setting, quantity
adjusters, the eect of the monetary tightening has its immediate eect
upon output and eventually upon unemployment. Any downward pressure
on prices occurs only subsequently as the result of prolonged underutilization of resources. And the slowing of monetary growth, so often misidentified in the orthodox literature as the cause of falling inflation, occurs only
because firms falling production requires less working capital in the form
of bank loans.
Interestingly, the Post Keynesian definition of monetary policy and its
analysis of the transmission mechanism is not far removed from that
adopted by mainstream economists with experience of the realities of
central banking (Goodhart 1994) or from the statements of central banks
themselves (Bank of England 1999). There are, though, major dierences
on the goals of monetary policy.
Given the belief that there are no automatic mechanisms for ensuring full
employment and their tendency to support Keyness proposition that it is
more important to disappoint the rentier than the worker, there is a natural
preference among Post Keynesians for a monetary (interest rate) policy that
takes full employment as its primary objective. This is flatly at odds with
the objectives of monetary policy as conducted by governments and central
banks in developed countries for the last twenty years or so. In practice, the
emphasis has been upon price stability (or at least low inflation) as the main
or even only goal of monetary policy, achieved by sustained high real interest rates accompanied by chronically depressed output. Intellectually, this
has drawn support from the rational expectations and policy irrelevance
work of the 1970s, which argued that there is no sacrifice involved (in
output or employment) since monetary policy can only aect prices.
Minimizing inflation is a free lunch. Such arguments amount to a restatement of the classical dichotomy wherein money is a veil whose quantity
determines the overall level of prices but has no eect upon quantities
which are determined by real forces. In its modern incarnation, this commitment to the classical dichotomy takes the form of inflation targeting
Monetary policy 259

wherein interest rate policy is assigned the sole function of minimizing
medium-term inflation and real magnitudes enter the picture only as indicators of likely future movements in inflation. This can be seen most clearly
where central banks adopt a Taylor-type rule for the setting of interest
rates. Measurements of the output gap are important, but only because
they provide information which can be used in the choice of interest rate
whose sole purpose remains to minimize inflation. If central banks adhere
rigidly to a Taylor rule and the object of monetary policy is solely the minimizing of inflation, then we are not far removed from the rule-governed
monetary policy of the kind advocated by Milton Friedman and others in
the 1960s. The interest rate replaces the stock of money as the instrument
of policy and the instrument is set according to a rule which features output
and real interest rates rather than the long-run growth of output.
To their credit, some mainstream economists (Laidler 1990; Blanchard
et al. 1997) have consistently rejected this view, faced with clear evidence
that the tightening of monetary policy in the early 1980s and 1990s clearly
played a part in the accompanying recessions. More importantly, the
general public has always been sceptical of any argument that monetary
policy does not aect their jobs and incomes. This has forced governments
to recognize the danger that operating a tight monetary policy was likely to
lead to electoral defeat if they were seen as authors of a policy in which the
public perceived real costs. From this was born the belief that monetary
policy was best operated by an independent central bank which did not
have to confront democratic preferences. The result has been a creeping privatization of monetary policy in which central banks have been reconstituted with charters giving them at least instrument independence (for
example, the Bank of England in 1997) and in some cases the freedom to
set the inflation target as well (the European Central Bank in 1999).
In the Post Keynesian view, therefore, monetary policy as operated has
helped sustain an unwarrantedly high level of unemployment and loss of
output. Furthermore, because monetary policy has become a one-club,
interest rate, policy it also has distributional eects which work against
broadly the same groups as are disadvantaged by chronic restrictions of
output. The distributional eects of interest rates operate most obviously
through agents portfolios, in particular through their net holdings of floating rate assets. Where these holdings are positive, a rise in interest rates
redistributes income in their favour (abstracting from any eects of the
local tax regime). By the same mechanism, those with net floating rate
liabilities lose. Thorbecke (1999) has shown that in periods of disinflationary policy the burden of adjustment is unequal between production sectors
(construction and durables suer most) and between social groups (lowincome workers and minorities pay the highest price).
260
Monetary policy
At the heart of Post Keynesian monetary policy, therefore, is not so

much a body of technical analysis which cuts it o from the mainstream (or
at least from its more realistic practitioners) but a desire to rid the practice
of policy from its deflationary biases, to reassert the value of discretion in
responding to monetary shocks and to restore accountability in the
conduct of monetary policy. At the heart of Post Keynesian policy is lower
interest rates. But it is widely recognized that such a policy will face severe
constraints, especially in the form of reactions from global financial
markets. Much discussion of what a Post Keynesian monetary policy might
look like in practice is thus taken up with suggestions for intervention in
and regulation of the financial system. For countries with independent
central banks enjoying goal and instrument independence, legislation may
be required to force a change in targets towards less-deflationary policy. For
central banks with only instrument independence, governments have
retained the power to set less-disinflationary targets. Any unilateral moves
of this kind, however, especially if accompanied by a public commitment
to a permanent shift in policy goals, would immediately bring an adverse
reaction from foreign exchange markets. Post Keynesian monetary policy
recognizes therefore the need for the redesign of the international monetary
system aimed at reducing the scale and volatility of foreign exchange flows
and creating a means whereby deficit countries can adjust without necessarily reducing demand. Some form of adjustable peg system seems most
appropriate, and suggestions have been made by Davidson (1992). Lowering interest rates has the inevitable consequence of giving an immediate
boost to the present value of all (including financial) assets as well as
increasing the demand for credit, much of which may later find its way into
asset-price inflation rather than productive use. Such dangers could be
overcome by a willingness to consider lending ratios imposed either upon
financial institutions who could be encouraged by the appropriate use of
such ratios to favour certain types of lending, or by ratios (of loan to spending) imposed upon borrowers in the way that hire-purchase agreements
once used to specify minimum deposits (Arestis and Sawyer 1998).
P H
See also:
Central Banks; Economic Policy; Endogenous Money; Inflation; New Classical Economics;
Rate of Interest.
References
Arestis, P. and M.C. Sawyer (1998), Keynesian economic policies for the new millennium,
Bank of England (1999), The transmission mechanism of monetary policy, Bank of England
Quarterly Bulletin, May, 16170.
Money 261
Blanchard, O., A. Blinder, M. Eichenbaum, R. Solow and J.B. Taylor (1997), Is there a core
of practical macroeconomics that we should all believe?, American Economic Review, 87
(2), 23046
Davidson, P. (1992), Reforming the worlds money, Journal of Post Keynesian Economics 15
(2), 15379.
Goodhart, C.A.E. (1994), What should central banks do? What should be their macroeconomic objectives and operations?, Economic Journal, 104 (427), 142436.
Laidler, D.E.W. (1990), Taking Money Seriously, Hemel Hempstead: Philip Allan.
Moore, B.J. (1988), Horizontalists and Verticalists, Cambridge: Cambridge University Press.
Thorbecke, W. (1999), Further evidence on the distributional eects of disinflationary monetary policy, Levy Institute Working Paper No. 264.
Money
Defining money is a vexing problem for monetary theorists. Readers are
familiar with the two usual approaches defining money by its functions,
or simply and arbitrarily choosing some empirical definition (as Keynes did
in the General Theory: we can draw the line between money and debts
at whatever point is most convenient (Keynes 1936 [1964], p. 167).
However in the Post Keynesian approach, the critical distinction is between
a unit of account and a thing that is denominated in a unit of account (following Keynes of the Treatise: the money-of-account is the description or
title and the money is the thing which answers to the description (Keynes
1930 [1976], p. 3). Many theorists make no such distinction, as they use the
term to sometimes refer to the thing (the medium of exchange) and other
times to refer to the title. To avoid confusion, I shall carefully distinguish
among money (the title, or dollar in the US), high-powered money (a particular money-thing reserves and currency), and bank money (another
money-thing demand deposits or private bank notes).
In the Post Keynesian view, money is not simply a handy numraire in
which prices, debts and contracts happen to be denominated. This contrasts with a general equilibrium approach, in which we may choose any
one good to serve as numraire, converting relative values to nominal
values in terms of the numraire. Indeed, the typical story of the origin of
money is really based on a numraire approach, in which Robinson Crusoe
decides to use tobacco, leather and furs, olive oil, beer or spirits, slaves or
wives . . . huge rocks and landmarks, and cigarette butts as money
(Samuelson 1973, pp. 2756). When, say, seashells are chosen as money by
Crusoe, he has simultaneously chosen a numraire and designated which
commodity money will serve as the money-thing. Eventually, Crusoe discovers that gold again, both a numraire and a money-thing has superior properties.
The conjectural history propagated by Samuelson (and many others) is
262
Money
dismissed by all serious historians and anthropologists. Interested readers

are referred to numerous accounts that emphasize the social nature of the
origins of money (see Wray 1998 for extensive references to the literature).
In any case, our primary purpose for examining history and pseudo-history
is to illuminate the nature of modern money. In my view, a system based on
a commodity or numraire money is not a money economy as Keynes
defined it. Rather, an economy in which money serves as nothing more than
a numraire is what Keynes called a barter or real wage economy. Even if
there really has been a historical stage in which there was a commodity
money, I would argue that it sheds no light on the operation of our modern
money system, in which both the unit of account and the money-things
denominated in that account arise from social practices (Ingham 2000).
Thus Post Keynesians emphasize the dierence between two approaches
what Goodhart (1998) has called C-form (Chartalist) and M-form
(Metallist, or commodity money) that is to say, between a theory in which
money is a social unit of account or that in which money is nothing more
than a numraire commodity adopted for convenience. The C-form
approach (or what has also been called the state money, or taxes-drivemoney, theory) insists, as did Keynes, that the state writes the dictionary
(decides what will be the money of account for example, the dollar in the
US) in all modern economies. This goes a long way towards explaining
what would appear to be an otherwise extraordinary coincidence: the onenation-one-currency rule. As Nobel laureate Robert Mundells work makes
clear, if money is simply a numraire chosen to facilitate exchange, then one
would expect use of a particular numraire within an optimal currency
area (Goodhart 1998). There is no reason to expect such to be coincident
with nation states. In fact, however, the one-nation-one-currency rule is
violated so rarely that exceptions border on insignificance, and those few
cases are easily explained away as special cases, as Goodhart demonstrates.
The European Union thus represents a substantial and perhaps risky
exception.
This leads us to an explanation of the use of money: why is money used?
The orthodox story begins, as we have seen, with Crusoe and Friday who
grow tired of the inconveniences of barter. In any case, money comes out
of the market. An alternative view that is consistent with a social approach
to money argues that money derived from the pre-civilized practice of
wergeld; or to put it more simply, money originated not from a pre-money
market system but rather from the penal system (Goodhart 1998; Wray
1998). An elaborate system of fines for transgressions was developed and,
over time, authorities transformed this system of fines paid to victims for
crimes to a system that generated payments to the state. Until recently, fines
made up a large part of the revenues of all states. Gradually, fees and taxes
Money 263
as well as rents and interest were added to the list of payments that had to
be made to authority.
According to the C-form or taxes-drive-money approach, the state (or
any other authority able to impose an obligation whether that authority
is autocratic, democratic or divine) imposes an obligation in the form of
a generalized, social unit of account: a money. This does not require the
pre-existence of markets, or of a numraire, or of prices of any sort. Once
the authorities can levy such an obligation, they can then name exactly
what can be delivered to fulfil this obligation. They do this by denominating those things that can be delivered, in other words, by pricing them in
the money unit.
Thus far we have only explained the money of account (the description).
Once the state has named the unit of account, and imposed an obligation
in that account, it is free to choose the thing that answers to the description. The state-money stage reaches full development when the state actually issues the money-thing answering to the description it has provided
that is, high-powered money. Economists often distinguish between a commodity money (say, a full-bodied gold coin) and a fiat paper money.
However, regardless of the material from which the money-thing issued by
the state is produced, the state must announce its value.
Indeed, in spite of the amount of ink spilled about the Gold Standard,
it was actually in place for only a relatively brief instant. Throughout most
of history, the money-thing issued by the authorities was not gold-money
nor was there any promise to convert the money-thing to gold (or any other
valuable commodity). It should be noted that for most of Europes history,
the money-thing issued by the state was the hazelwood tally stick. Other
government-issued money-things included clay tablets, leather and base
metal coins, and paper notes. Why would the population accept otherwise
worthless sticks, clay, base metal, leather or paper? Because the state
agreed to accept the same worthless items in payment of obligations (fees,
fines and taxes) to the state.
Georg Friedrich Knapp (18421926) distinguished between definitive
money accepted by the state in (epicentric) payments of obligations to the
state, and valuta money used by the state in its own payments (apocentric) (Wray 1998). In todays modern money systems, high-powered money
fulfils both functions. Of course, it appears that the US government accepts
bank money in payment of taxes, but in reality payment of taxes by bank
cheque leads to a reserve drain from the banking system. Government
spending, of course, takes the form of a Treasury cheque, which when
deposited in a private bank leads to a reserve credit. Note that, so long as
government does accept bank money in epicentric payments at par with
high-powered money, from the point of view of the non-bank public there
264
Money
is no essential dierence between bank money and high-powered money.

This is not true for banks, which lose reserves when taxes are paid by bank
cheque and gain reserves whenever Treasury cheques clear.
Finally, Knapp defined as paracentric those payments made between
non-government entities. In all modern economies, these mostly involve use
of bank money and other money-things issued by the non-government
sector (what can be called inside or credit money). There is a hierarchy
or pyramid of money-things, with non-banks mostly using bank moneys
for net clearing and with banks using high-powered money for net clearing
with other banks and with the government. Note that all these moneythings are denominated in the unit of account, that is, the account in which
obligations to the state are enumerated, and all credit money-things also
represent a social relation that between creditor and debtor.
Post Keynesians are best known for their work on credit money and the
endogenous money approach. Because there is a separate entry on endogenous money in this volume, we need only briefly summarize the endogenous money approach, while explaining the relation between credit money
and state money.
The evolving Post Keynesian endogenous approach to money oers a
clear alternative to the orthodox neoclassical approach that is based on
central bank control of an exogenous money supply through provision of
reserves. Early Post Keynesian work emphasized uncertainty and was generally most concerned with hoards of money-things held to reduce disquietude, rather than with money-things on the wing (the relation with
spending). However, Post Keynesians always recognized the important role
played by money in the monetary theory of production that Keynes
adopted from Marx. Circuit theory, mostly developed in France, focused
on the role money plays in financing spending. The next major development came in the 1970s, with Basil Moores horizontalism (somewhat anticipated by Nicholas Kaldor), which emphasized that central banks cannot
control bank reserves in a discretionary manner. Reserves must be horizontal, supplied on demand at the overnight bank rate (fed funds rate)
administered by the central bank. This also turns the textbook deposit
multiplier on its head, as causation must run from loans to deposits and
then to reserves.
This led directly to development of the endogenous money approach
that was already apparent in the circuit and Marxist literature. When the
demand for loans increases, banks normally make more loans and create
more banking deposits (money-things), without worrying about the quantity of reserves on hand. Privately-created credit can thus be thought of as
a horizontal leveraging of reserves, although there is no fixed leverage
ratio.
Multiplier
265
Like Keynes, Post Keynesians have long emphasized that unemployment

in capitalist economies has to do with the fact that these are monetary economies. Keynes had argued that the fetish for liquidity (the desire to hoard)
causes unemployment because it keeps the relevant interest rates too high
to permit sucient investment. While it would appear that monetary policy
could eliminate unemployment, either by reducing overnight interest rates
or by expanding the quantity of reserves, neither avenue will actually work.
When liquidity preference is high, there may be no rate of interest that will
induce investment in illiquid capital and even if the overnight interest rate
falls, this may not lower the long-term rate. Further, as the horizontalists
make clear, the central bank cannot simply increase reserves in a discretionary manner as this would only result in excess reserve holdings, pushing
the overnight interest rate to zero without actually increasing the supply
of private money-things. Indeed, when liquidity preference is high, the
demand for, as well as the supply of, loans collapses. Hence there is no way
for the central bank to simply increase the supply of money in order to
raise aggregate demand. This is why those who adopt the endogenous
money approach reject ISLM analysis in which the authorities can eliminate recession simply by expanding the money supply and shifting the LM
curve out.
L. R W
See also:
Banking; Central Banks; Circuit Theory; Endogenous Money; Liquidity Preference;
Monetary Policy.
References
Goodhart, C. (1998), The two concepts of money: implications for the analysis of optimal
currency areas, European Journal of Political Economy, 14 (3), 40732.
Ingham, G. (2000), Babylonian madness: on the historical and sociological origins of
money, in J. Smithin (ed.), What is Money?, London and New York: Routledge, pp. 1641.
Keynes, J.M. (1930 [1976]), A Treatise on Money. Volume 1: The Pure Theory of Money, New
York: Harcourt Brace Jovanovich.
York: Harcourt Brace Jovanovich.
Samuelson, Paul (1973), Economics, 9th edition, New York: McGraw-Hill.
Wray, L. Randall (1998), Understanding Modern Money: The Key to Full Employment and
Price Stability, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Multiplier
A key aspect of Keyness impact upon political economy has been the role
of the multiplier in justifying injections of government expenditure into
an unemployment-prone capitalist economy. For Keynes, writing in the
266
Multiplier
depression of the 1930s, the possibility that each pound of Treasury expenditure could generate a multiplied eect on total income represented a valuable political argument. Borrowing from the work of his student, R.F.
Kahn, Keynes was able to argue that each newly employed worker would
carry out expenditure that resulted in a ripple eect of additional employment and income, the multiplier measuring the size of the total expansion.
Moreover, this impact on total income would even generate sucient savings
to fund the initial injection of expenditure, with no cost to the Treasury.
To illustrate this argument, consider an initial injection of investment
(I) by firms. This injection generates new income in a series of rounds. In
the first round there is an increase in income (I), out of which expenditure
increases according to the propensity to consume b. Hence in the second
round income increases by bI. In the third round this newly generated
income generates a further bout of expenditure resulting in an increment
b(bI). This process will continue until eventually the impacts peter out. The
overall impact upon income is:
Y(1bb2 b3 . . .)I.
(1)
Taking the sum of the polynomial contained in the brackets:

Y
1
I.
1b
(2)
This is the multiplier relationship between increments in investment and

income. So long as the propensity to consume b is less than 1, the multiplier
takes a value more than 1. More income is generated than the initial injection in investment; the impact upon the economy is multiplied, hence the
term multiplier.
Since the propensity to consume defines the proportion of income that
is spent on goods and services, 1 b is the propensity to save. It follows that
(1 b)Y represents the total volume of savings (S) generated by the
injection of investment. Taking the denominator of the multiplier equation
to the left-hand side we have the identity:
SI.
(3)
This is the second politically attractive feature of the multiplier model: the
injection of investment is financed out of savings generated by the incomegenerating process. Not only does investment generate more income than
the initial outlay; it is also self-financing. There is a compelling argument,
under circumstances of high unemployment, for the state to intervene to
ensure that income is stimulated by new investment.
Multiplier
267
For Post Keynesians this analysis is fraught with diculties because of

the time that it takes for new savings to be generated by the multiplier
process. New savings only match the initial injection of investment after a
series of rounds in which consumer spending reacts to changes in income.
Indeed, it has been argued that firms must borrow money in order to
finance investment in the hope that savings will be forthcoming that can be
used to pay o their debts, as the multiplier process works itself out. Since
the multiplier process is not instantaneous, it is necessary to consider the
way in which investment is financed.
The main source of finance is the banking system. Consider what
happens when a bank agrees to make a loan for a new investment project.
A remarkable institutional observation, which is now central to much of
Post Keynesian thought, is that this loan is both an asset and a liability.
After the loan is granted it is used by the firm to pay the supplier of new
equipment and to hire new workers. The recipients now hold this outlay as
deposits in the banking system. The initial loan therefore represents an
asset (a promise by the firm to pay) and a liability (a promise by the banking
system to pay).
For some Post Keynesians, this institutional observation has led to the
conclusion that the multiplier process is not required as an explanation of
how investment is financed. Since each loan is a deposit, the banking
system is capable of financing investment without relying on a multiplier
process to generate more income and savings. Indeed, for Moore (1994) an
injection of investment is instantaneously equal to savings because of the
deposits generated by each new loan.
A similar position is adopted in the French circulation approach. Money
is viewed as changing hands in a closed circuit, from banks to firms and
households, and back to the banking system. The problem with the multiplier process is that changes in investment generate changes in income, but
implicitly there is a dead weight of total income that remains unchanged.
For Schmitt: If some incomes are created how can we explain that other
incomes are simply maintained in inertia through time, where they are
deemed to be neither created nor destroyed? (Deleplace and Nell 1996,
p. 125). In the circulation approach all income is generated by injections
from the banking system that return back to the banking system as part of
the money circuit. The multiplier is an obstacle to seeing clearly the conditions under which circuits are completed.
The importance of the multiplier can be defended, however, by arguing
that even though the multiplier process is not a required condition for the
finance of investment, its impacts should still be taken into account. A
change in investment will still result in more workers being employed,
and those workers will spend additional income on additional goods and
268
Multiplier
services. Indeed by ignoring the multiplier it is possible to overestimate the

amount of money that needs to be advanced by the banking sector in order
to enable a complete circuit of money. Without multiplier eects that ripple
between firms and households, banks are required to advance all of the
money required for both consumption and investment purposes at the start
of the circuit. In circuit theory the web of debt may appear larger and hence
more prone to instability than is actually the case once the multiplier
process is taken into account.
Although the multiplier has been considered thus far as a dynamic
process, Chick (1983) demonstrates that in Keyness General Theory this
dynamic approach is defined alongside a comparative static interpretation.
In the static variant there is a precise focus comparing dierent points of
static equilibrium. Consider a closed economy in which income (Y) is made
up of consumption (C) and investment (I):
YCI.
(4)
Assuming that consumption is dependent upon income such that CbY,

with b representing the propensity to consume:
YbYI
(5)
which by manipulation can be written as

Y
1
I.
1b
(6)
As before, the term 1/1b is the multiplier, but here the relationship
between aggregate income and investment is specified instead of the relationship between changes in these magnitudes. Paul Davidson argues that in
comparative statics the multiplier should be used to compare two economic
systems, each with identical propensities to consume, but dierent volumes
of investment. Following an earlier insight provided by Joan Robinson, this
can be viewed as a controlled experiment in which two systems are compared holding everything else constant (Davidson 1994, pp. 4041).
To some extent this interpretation limits the use that can be made of the
multiplier. The multiplier is not used here as a method of explaining the
actual course taken by a capitalist economy over time. The points of equilibrium can be compared but, in contrast to multiplier process analysis,
there is no attempt to explain how an economy moves between two positions of equilibrium. As a consequence, Davidson contends that the multiplier is marginal to Keyness system.
Multiplier
269
Moreover, Davidson argues that in the early 1930s Keynes had worked
out the substance of the General Theory without the multiplier, only adding
it in for political reasons. This contrasts directly with Patinkins revealing
quotation of Keynes: The essential role that the multiplier plays in the
General Theory is attested to by Keynes declaration to Beveridge, shortly
after its publication, that about half the book is really about it (Patinkin
1982, p. 199). Patinkin (ibid., p. 19) also relies on the multiplier to dismiss
Michal- Kaleckis claim to have discovered the substance of the General
Theory prior to Keynes, arguing that this would be impossible as Kalecki
did not discover the role of the multiplier in his system until 1943.
Aside from discussion about the origins of the General Theory, a case can
be made for the analytical power provided by the multiplier in establishing
the conditions that are required for full employment. Stated simply, the
principle of eective demand shows that firms will not in general produce
at full employment because of the leakage of savings from the economic
system. As Chick (1983, p. 253) makes clear, the multiplier equation (6)
says exactly what the Principle of Eective Demand says: that for a given
level of income to be sustainable, the gap between income and consumption must be filled with investment. Since consumers have positive savings,
only a part of income is realized by consumer demand. The shortfall must
be taken up by investment, and Keynes shows that in general private investment cannot be expected to do this job.
The importance of the multiplier in specifying the conditions required
for full employment is testified by its relevance to long-run analysis. On the
boundaries of Post Keynesian theory it has been argued that the principle
of eective demand is only truly general in its applicability when extended
to the long run. The key problem with Keyness short-run period of analysis is that investment is viewed only as a component of aggregate demand.
There is virtually no attention paid to the increase in productive capacity
that will result in the next period. In the HarrodDomar growth model the
multiplier has a key role in identifying the necessary conditions for matching aggregate demand with productive capacity in the long run. Domar
(1957) develops a dual approach to investment, one side of which is its
capacity-generating role, the other its impact on demand via the multiplier.
The requirements on aggregate demand to match this growth in capacity
render the maintenance of full employment even more unlikely than is perceived in a short-run framework.
Long-run analysis also provides the basis for interaction with corresponding traditions in economic thought. Domar identified the close relationship
between his growth model and Marxs reproduction schema, with the multiplier taking centre stage (Trigg 2002). Similarly, the multiplier is central to
an emerging Sraan literature on the relationship between eective demand
270
Multiplier
and long-run capacity utilization (Serrano 1995). Common to both this and
the Marxian approach is the observation that a monetary production
economy must necessarily also be a surplus-producing economy. Investment
goods that allow an expansion of productive capacity are by definition
surplus goods. It should be noted, however, that in order to facilitate a
simple macroeconomic exposition, this literature is limited to the narrow
assumption that only one commodity is produced. Since monetary economies are necessarily multisectoral, with money providing the mechanism for
exchanging heterogeneous commodities, the issue of aggregation requires
close methodological attention a challenge that applies with equal force to
the models of circuit theory and multiplier process analysis.
A B. T
See also:
Circuit Theory; Consumption; Eective Demand; Endogenous Money; Finance Motive;
Fiscal Policy; Growth Theory; Keyness General Theory.
References
Chick, V. (1983), Macroeconomics After Keynes, Oxford: Philip Allan.
Davidson, P. (1994), Post Keynesian Macroeconomic Theory, Aldershot, UK: Edward Elgar.
Deleplace, G. and E.J. Nell (eds) (1996), Money in Motion: The Post Keynesian and Circulation
Approaches, Basingstoke: Macmillan.
Domar, E.S. (1957), Expansion and employment, in Domar, Essays in the Theory of
Economic Growth, New York: Oxford University Press, pp. 83108.
Moore, B.J. (1994), The demise of the Keynesian multiplier: a reply to Cottrell, Journal of
Post Keynesian Economics, 17 (1), 12134.
Patinkin, D. (1982), Anticipations of the General Theory?, Oxford: Basil Blackwell.
Serrano, F. (1995), Long period eective demand and the Sraan supermultiplier,
Contributions to Political Economy, 14, 6790.
Trigg, A.B. (2002), Surplus value and the Keynesian multiplier, Review of Radical Political
Economics, 34 (1), 5767.
New Classical Economics

New Classical economics is a macroeconomic doctrine designed to oppose
Keynesian policies and theories, and demonstrate the self-regulating powers
of the capitalist system. Its leaders included Robert Lucas, Thomas Sargent,
Neil Wallace and Robert Barro. The doctrine flourished especially in the US
and during the 197585 decade when most of its notable works were written.
It has since sunk into decline because the circumstances that gave rise to the
doctrine have changed, but it is fair speculation to say that it may manifest
itself again, at some appropriate time and in another form.
The classical economists in the early nineteenth century had argued that
capitalism was a self-regulating system which would not, by itself, generate
recessions and a business cycle. A century later J.M. Keynes replied that the
classical economists had forgotten uncertainty, which could cause a drive for
liquid assets at the expense of commodity demand. Keynes therefore found
inherent fault in the capitalist system, and argued that the economy needed
regulation at the macro level. The New Classical economists replied that,
since microeconomic theory did not recognize uncertainty, nor should macro
theory. They believed that rational behaviour meaning maximization
should be the dominant theme of both micro and macro theory, from which
they concluded that Keynesian macroeconomic policies were ineective.
Keynes had said that classical theory did not explain the economy, and
the New Classical economists replied with a point about what constituted
an acceptable theory. The New Classical innovations were primarily of a
methodological nature, but their formulators were theoreticians rather than
philosophers, and their innovations were presented in sophisticated mathematical terms and as scientific discoveries.
New Classical references to the old classical predecessors are rare and
sparse, and it is obvious that they were more directly influenced by
Monetarism, except on one particular point where that doctrine seemed
inconsistent with reason and experience. In 1968 Milton Friedman had distinguished between anticipated and unanticipated monetary changes, in
order to argue that countercyclical policy would normally have no lasting
beneficial eects. His argument was that if the financial markets thought in
real terms, and took prospective inflation into account, then nominal
changes (in the money supply) would lead only to nominal changes (in the
price level). Keynesians had argued that an increase in the money supply
would have real eects, in that it would reduce the rate of interest;
Friedman replied that in a rational economy the real rate of interest would
271
272
New Classical economics
remain unchanged, because monetary increases would simply lead to inflation. Friedman concluded that the Phillips curve operated only in the short
run, and that there was a natural rate of unemployment that could not be
changed by Keynesian policy.
Friedmans attack on Keynesian demand management precipitated the
widespread liberalization and deregulation of markets during the last
quarter of the twentieth century. Many of these changes were liberating,
but some were potentially destabilizing, and required a high degree of confidence in the self-regulating powers of the macro economy. Yet Friedmans
rebuttal of Keynes depended on a particular and dubious assumption.
Specifically Friedman assumed that the velocity of money would remain
constant regardless of other economic change. For if monetary velocity did
remain constant then increases in the money supply would have only inflationary eects, and tendencies to hoard would have only deflationary
eects, without changing the level of unemployment.
Yet the evidence seemed to show that monetary velocity increased during
booms and decreased during recessions, in response to changes in the precautionary demand for money. There also seemed to be a long-run tendency to economize on the use of money, in accordance with the growing
sophistication of the capital markets. Finally, the constant velocity of
money proposition seemed arbitrary because it did not follow from microeconomic theory. Friedman advanced econometric evidence for the constant velocity proposition, and he also argued that what mattered was the
results and not the theory and its assumptions. Nevertheless, a whole
philosophy of government could hardly be based on an empirical relationship that was inexplicable and at best dicult to discern.
New Classical theory rose to prominence by arguing that the refutation
of Keynes did not require a constant velocity of money, because the selfequilibrating quality of the capitalist system followed from pure microeconomic theory. The only assumption required was that individuals would act
rationally as elementary microeconomic theory assumes. The stability of
capitalism, and the inecacy of countercyclical policy, could then be demonstrated as a matter of mathematics.
This theme of rational behaviour was the major innovation of New
Classical economics; and most of Friedmans conceptual innovations,
including the natural rate of unemployment and the distinction between
anticipated and unanticipated changes in the money supply, became corollaries that followed from it. If decision makers rationally maximized then
Says Law would hold, markets would clear, and capitalist stability would
follow.
Rationality also meant that Keynesian policies would have no eect. (For
some mysterious reason the following argument is known as the Ricardian
New Classical economics 273

equivalence theorem.) Keynesian theory postulated that government
budget deficits would stimulate the economy and increase employment.
The New Classicals conceded that government spending, by itself, was
stimulatory. However, government budget deficits also carried the prospect
of future tax rises; and rational markets would understand that the capitalized value of the implied future taxes would approximately equal the value
of the additional government spending. The prospect of these future taxes
would reduce spending, by just as much as the government had increased
it. Therefore what the Keynesians had thought was an expansionary fiscal
policy was not expansionary at all.
Keynesian theory had assumed only implicitly but nevertheless crucially that the financial markets were subject to money illusion. As a
matter of fact, basic Keynesian theory demonstrated that there would be
inflation after the money supply increased. However, Keynesian theory
implicitly assumed that market agents did not know the theory because if
the markets were rational, nominal changes would have only nominal
eects. The Keynesian theory of knowledge made no sense, because it
assumed that the policy makers knew the theory and that the agents did not.
Yet it was impossible to modify the theory, because dual knowledge rational policy makers but money illusion in the markets was fundamental.
The main question was, what did the markets really know? The New
Classical answer, which had been given by John Muth, was that the markets
would economize information while making maximizing choices. They
would learn over time, using the available information in the most ecient
way. Rational markets would learn how to respond most eectively to
Keynesian policies or any other exogenous changes that would disturb the
system. Thus, if a central bank adjusted the money supply, the adjustment
would have no real eect on a rational market system unless the change
was a signal that the bank knew something that the markets did not. Thus
Muth writes:
I would like to suggest that expectations, since they are inferred predictions of
future events, are essentially the same as the predictions of the relevant economic
theory . . . The [rational expectations] hypothesis asserts three things: (i) information is scarce, and the economic system generally does not waste it; (ii) the
way expectations are formed depends specifically on the structure of the relevant
system describing the economy; (iii) a public prediction will have no substantial eect on the operation of the economic system unless it is based on insider
information. (Muth 1961, p. 316)
A discernible air of excitement permeates many of the New Classical

writers, who must have felt that they were at the very conjunction of transforming ideas. There was a new macroeconomic doctrine that flowed elegantly from microeconomics, and for the first time integrated micro and
274
New Classical economics
macro theories. The social and political implications included justification

for a greater degree of economic freedom. And their system seemed to
work, because from 1975 to 1985 most national economies were troubled
primarily on the supply side, and the demand-side problems that the
Keynesians had addressed were not on the horizon.
One major theoretical problem was not addressed. The theory was that
(except for a random variable) the market would be able to predict future
prices. This followed from elementary microeconomics, because rationality
maximization is only meaningful if there is full knowledge, or some stochastic version of full knowledge. Then, if the market can predict the
future, countercyclical policies will have no eect for the reasons given
above. However maximization implies that economic behaviour is consistent and in principle predictable.
If there is uncertainty about the future, then, microeconomic theory notwithstanding, market behaviour will not be predictable and consistent.
When Muth noted that rational people do not waste information, he should
also have said that people cannot act rationally in the sense of New
Classical economics unless they have a high degree of information.
Maximization is impossible unless the constraint is understood, or in other
words it is necessary to know the opportunity line as well as the indierence curve. For, when information is poor, and economic agents do not
understand their economic constraints, the animal spirits and states of confidence that are emphasized by the Keynesians can dominate the macroeconomic picture. Economic agents may advance their self-interests in an
intelligent way, using what information they have as fluidly as possible, but
there is no reason why their actions should be consistent over time.
Uncertainty does not cancel out, but renders economic behaviour fluid and
indeterminate.
If there were a small increase in the money supply in a stable economy
then it would be reasonable to expect that prices would increase in proportion to the increase in the money supply. In such circumstances, where there
is a high degree of information, the Keynesians were wrong and New
Classical theory comes into its own. But in more general conditions, when
people do not more or less know what the future will bring, and economic
agents cannot optimize, Keynesian policies constitute eective signals.
The eectiveness of Keynesian policies is the norm, and rationality in the
New Classical sense is the exception, because macroeconomic instability
arises in circumstances of imperfect information.
New Classical economics became irrelevant when it was overtaken by the
course of events. From the mid-1980s to the early years of the twenty-first
century, the global economy went from boom to recession to boom again
and recession again. There is nothing in New Classical theory to explain
New Keynesian economics
275
this sequence of destabilizing swings in aggregate demand, nor is there anything that would suggest a remedy. The main ideas of Monetarism and New
Classical theory now live on in New Keynesian economics, which is another
supply-side doctrine that recognizes rational behaviour and the natural
rate of unemployment. The dierence is that New Keynesian economics
does concede that there are faults and diculties in the economy, and it also
recognizes the possible mismatch of aggregate demand and supply.
However it is a compromise doctrine that lacks the logical rigour of New
Classical theory.
The New Classical economists raised important issues of method and
theory. They demonstrated some important faults and errors in the
Keynesian system, and their ideas encouraged the process of economic
deregulation at a time when Monetarism seemed to be failing its own
empirical tests.
A F
See also:
Agency; Economic Policy; Expectations; Fiscal Policy; Inflation; Monetary Policy; Nonergodicity; Says Law; Uncertainty; Walrasian Economics.
Bibliography
Fitzgibbons, A. (2000), The Nature of Macroeconomics: Instability and Change in the
Capitalist System, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Friedman, M. (1968), The role of monetary policy, American Economic Review, 58 (1), 117.
Hoover, K. (ed.) (1992), The New Classical Macroeconomics, two vols, Aldershot: Edward
Elgar.
Lucas, R. and T. Sargent (1981), Rational Expectations and Econometric Practice, Minnesota:
University of Minnesota Press.
Muth, J. (1961) Rational expectations and the theory of price movements, Econometrica, 29
(3), 31535.
New Keynesian Economics

Central to New Keynesian theory is the notion that wages and prices do not
adjust rapidly enough to achieve the self-regulation of classical and neoclassical economics. The latter theories assume frictionless markets, which
ensure rapid correction whenever the economy deviates from its long-run
equilibrium. New Keynesian theorists believe that market failures amplify
and lengthen such deviations, accounting for business cycles. Although it is
alleged to be a macroeconomic theory, its practitioners concentrate on
establishing the microeconomic foundations of the price and wage stickiness that is generated by market failures. The absence of market clearing is
supported by the argument that quantity rather than price adjustments are
276
in the interest of both workers and firms, that is, that market participants
behave rationally. In this way, New Keynesian economics purports to show
that decisions at the microeconomic level are optimal while capable of generating adverse eects at the macroeconomic level.
The claim that economic agents rational behaviour not merely prevents
the economy from absorbing shocks but amplifies and extends their eects
is in stark contrast to the New Classical view that rational behaviour ensures
a rapid return to equilibrium. Not surprisingly, the New Keynesian interventionist policy prescription creates a similar contrast. While agreeing that
in some cases private actions will oset public policy, they argue that this is
by no means always the case. They endorse built-in stabilizers and discretionary policy to reduce macroeconomic fluctuations, but are less optimistic about the power of policy than are Keynesians and Post Keynesians.
New Keynesians warn against discretionary policy that initiates osetting
private action, and reject fine-tuning as unrealistic. However, within these
parameters, they view government action as a remedy, rather than as a cause
of problems.
The beginning of New Keynesian theory is usually traced to work by
Fischer (1977) and Phelps and Taylor (1977), but studies of wage and price
inflexibilities as the result of rational behaviour substantially predate these
works and the use of the New Keynesian label. For example, J.R. Hicks
published work in the 1930s that stressed downward rigidity of nominal
wages as a key factor contributing to fair labouremployer relations that
enhance eciency. Okun (1975) considers fairness in cementing firms
relations with customers to explain price rigidities. Buyers regard price
increases in response to rising costs to be fair, but as gouging if they are a
response to rising demand. Other explanations of wage rigidity viewed
labour as a long-term investment in human capital (Doeringer and Piore
1971). Lastly, Keynes himself stressed the importance of relative wages in
explaining inflexibility. In all these examples, price and wage rigidities result
from rational behaviour.
New Keynesians distinguish between nominal and real rigidities. A
nominal rigidity prevents money prices from adjusting proportionately to
changes in nominal output. Real rigidities, such as the stickiness of a real
wage or of a relative wage or price, can be traced to firms acting to increase
eciency. Early New Keynesian eorts concentrated on discovering why
nominal wages are inflexible. Explanations included overlapping staggered
wage contracts and eciency wages. Overlapping contracts refers to the
real-world practice of labour and employers agreeing to wage contracts
that commonly cover periods ranging from one to three years. Moreover,
these contracts end at various times throughout the year. Consequently,
even though expectations may be revised as economic conditions change,
277
nominal wages are fixed in contracts that are due for renegotiation at intervals during the ensuing three years, delaying the adjustment indicated by
the revised expectations. Then, if demand falls, wage cuts are not possible,
leaving layos as the only option.
The macroeconomic costs of these quantity adjustments can be very
large, causing sometimes lengthy recessions while the wage adjustments
occur. Nevertheless, the long-term wage contracts that impose quantity
adjustments are preferred by both firms and workers. In the case of longterm contracts, the primary advantage is to reduce the high costs of negotiation borne by both firms and unions. These involve cash outlays as well
as time, not only for the negotiations, but also to conduct the necessary
research into existing wages and working conditions in comparable firms,
and into assessing the economic conditions inflation, employment, profits
likely to prevail over the life of the contract. A second advantage is the
reduced opportunity for strikes, which are also costly to both labour and
employers.
A second explanation of nominal wage rigidities attributes slow adjustment to the practice of paying eciency wages. Eciency wages involve the
payment of a premium over the nominal wage predicted by traditional
labour market analysis. The premium arises from the claimed dependence
of productivity on the wage paid. The higher wage is expected to reduce
slacking and absenteeism because workers believe they are treated well by
their employer, and enables firms to attract better-quality workers. In addition, it increases workers attachment to the firm, reducing costs associated
with high turnover. The result is an improvement in productivity sucient
to justify the higher wage. For example, faced with an adverse demand
shock, firms will not reduce the eciency wage, since this would lower productivity and increase costs. Instead, the rational firm will cut employment.
This emphasis on the labour market was a natural first step for research
in the Keynesian tradition. However, as long as profits are flexible, nominal
wage stickiness is not sucient to explain fluctuations of real output. For
example, if profits are allowed to fall, nominal prices could fall, reducing or
eliminating output fluctuations. The next phase of work emphasized price
stickiness as a necessary condition for changes in real output, and dealt
with the question of why nominal prices are not flexible enough to mirror
changes in nominal output. In contrast to the competitive model used by
New Classical theory, most New Keynesian analyses use a model of
monopolistic competition, giving firms some control over price. The essential feature of this model is that price is always above marginal cost, so that
firms are willing to sell more at the existing price when demand increases,
while a perfect competitor would not sell more unless the price rose.
However, following a change in demand, the model predicts adjustment of
278
both price and quantity by the rational, profit-maximizing firm. Clearly,

monopolistic competition alone cannot account for sticky prices. Some
modification is needed to reconcile rationality and price stickiness.
One such modification has been the development of models of staggered
price contracts (for example, Blanchard 1986) with strong parallels to the
staggered wage models referred to above. These investigate the eects of
staggered price setting for overlapping periods, starting from the premise
that the length of the period during which the price remains fixed depends
on the costs of adjustment. The staggering of price adjustments is attributed to firm-specific shocks occurring at dierent times. As with overlapping
wage contracts, adjustment to changing demand conditions is achieved
slowly and in a piecemeal fashion.
A core contribution to New Keynesian theory is the menu cost approach.
Menu costs are associated with changing prices in response to changing
demand, such as the cost of publishing new price lists and catalogues, and
changing price tags. Broader definitions of menu costs also include managerial time taken as well as cash expenditures made to establish the need for
change and to renegotiate contracts with suppliers. Faced with a drop in
demand for its product, conventional analysis simply predicts that the
profit-maximizing firm will reduce price. New Keynesian analysis suggests
that reducing price will incur high menu costs, so that profit may well be
maximized by keeping price constant (Akerlof and Yellen 1985). The rational firm will therefore cut output, not price. It is claimed that even small
menu costs can cause severe recessions (Mankiw 1985).
The dierent explanations of the absence of market clearing have each
been subjected to particular criticism, usually by critics who implicitly
accept the general framework of analysis. One such criticism of the menu
cost approach is that while it stresses the costs incurred when prices are
adjusted, it fails to address the costs attendant upon quantity adjustment.
These would include costs of shutting down equipment, storing or scrapping semi-finished products, renegotiating contracts with input suppliers
(or paying contractually agreed penalties) and, ultimately, reversing these
to restore output levels later. In short, they bear great similarity to the menu
costs of price adjustment. A second problem with this approach concerns
the failure to consider menu costs in a dynamic context. The menu costs of
a price reduction might wipe out profit for a period of time, after which
profit would recover. The implicit assumption is that the time period equals
or exceeds the period of depressed demand.
These are relatively minor criticisms compared with the general shortcomings of New Keynesian economics. Indeed, it can be criticized as being
very narrow in its focus, and neither Keynesian nor new. First, the analyses
of rigidity are typically framed in terms of a shock to a representative
279
firm, without regard for the variability of non-market-clearing responses

and their causes. However, empirical observation shows that periods of
price rigidity are of extremely variable length, with prices changing frequently in some industries, and seldom in others (Carlton 1986). Carlton
also found that small price changes (for example, less than 1 per cent) are
not uncommon, suggesting very low menu costs. His paper also considers
industry structure, the type of product and the nature of relations between
firms and customers, all of which have implications for pricing behaviour,
and none of which is addressed by New Keynesian economics. Second, virtually no eort is made to incorporate the propagation mechanisms that are
central to what are usually accepted as macrodynamic models. This neglect
is clearly consistent with the underlying assumption of New Keynesian literature that, imperfections notwithstanding, the economy is self-regulating
in the broader sense that it hovers round a macroeconomic equilibrium at
the NAIRU (non-accelerating inflation rate of unemployment). Given this
characteristic, New Keynesian models cannot be regarded as Keynesian;
they are special cases of the neoclassical model. Finally, as suggested above,
these ideas are not new, but reprise an older literature that treats wage and
price rigidities. The earlier literature is richer in content, if less rigorously
presented.
New Keynesian economics as a distinct school of thought was shortlived. Although much of the recent research in asymmetric information,
credit rationing and coordination problems might be classified as New
Keynesian, the term is rarely encountered in literature published since the
early 1990s. Instead New Keynesian thought has become part of the
broader field of research and has been reabsorbed into the current mainstream concern with the economics of imperfections.
W C
See also:
Credit Rationing; Economic Policy; Microfoundations; New Clasical Economics; Pricing and
Prices; Unemployment; Wages and Labour Markets.
References
Akerlof, George and Janet Yellen (1985), A near-rational model of the business cycle with
wage and price inertia, Quarterly Journal of Economics, 100 (5), Supplement, 82338.
Blanchard, Olivier J. (1986), The wage-price spiral, Quarterly Journal of Economics, 101 (3),
54365.
Carlton, Dennis W. (1986), The rigidity of prices, American Economic Review, 76 (4), 63758.
Doeringer, P. and M. Piore (1971), Internal Labor Markets and Manpower Analysis,
Lexington, MA: D.C. Heath.
Fischer, Stanley (1977), Long-term contracts, rational expectations, and the optimal money
supply rule, Journal of Political Economy, 85 (1), 191205.
Mankiw, N. Gregory (1985), Small menu costs and large business cycles: a macroeconomic
model of monopoly, Quarterly Journal of Economics, 100 (2), 52938.
280
Non-ergodicity
Okun, A. (1975), Inflation: its mechanics and welfare cost, Brookings Papers on Economic
Activity, 2, 351401.
Phelps, Edmund S. and John B. Taylor (1977), Stabilizing powers of monetary policy under
rational expectations, Journal of Political Economy, 85 (1), 16390.
Non-ergodicity
In response to the growing hegemony of the rational expectations revolution and the increasing complaint levelled against Post Keynesian economics that its concept of uncertainty had not been formalized or empirically
evaluated, Paul Davidson introduced the notion that the Post Keynesian
conception of uncertainty could be articulated with reference to a technical distinction between ergodic and non-ergodic processes. Building on
Samuelsons suggestion that economic knowledge about the future rested
on the axiom of ergodicity, Davidson (198283) argued that the Post
Keynesian conception of unknowledge was predicated on the rejection of
the universality of the ergodic axiom. He suggested that the Post Keynesian
view of time and discussion of the salience of uncertainty could be defined
with respect to the absence of governing ergodic processes what he
labelled non-ergodicity.
Ergodic theory has been explicitly developed in the theory of stochastic
processes although the term itself arises from statistical mechanics (see
Parry 1987 for a technical discussion). It refers to the property by which the
time and space averages that originate and are computed from any datagenerating process either coincide for a series of infinite observations, or
converge as the number of observations increases (with a probability of
one) for a finite number. That is to say, averages from past realizations collapse on the objective probability distribution that governs current and
future outcomes. Under such conditions the past reveals the future, and the
rational expectations hypothesis that the process of competition forces
agents to use all the amount of available information in forming expectations about the future which are ecient, unbiased and without persistent
errors, appears reasonable in the long run at least (see Table 1, below).
In Davidsons (198283) seminal paper he argued that the rational expectations hypothesis was a misleading caricature of Keyness recognition of
the importance of uncertainty and expectations, because the ergodic
assumption implies that the past reveals the future that over time agents
can predict the future with actuarial certainty-equivalence. On the ergodic
hypothesis, the passage of time does not aect the joint probability laws
governing processes; history ultimately does not matter. In contrast,
Davidson argued that agents would be truly uncertain under conditions of
non-ergodicity, that is, in the absence of governing ergodic processes. This
Non-ergodicity 281
is not to deny a priori that some economic processes may be ergodic, at least
for short periods of calendar time. But, under non-ergodic conditions, sampling from the past in the manner implied by the rational expectations theorists is not sensible since, even if agents have the ability to assemble and
process all the relevant information pertaining to past and present outcomes, the future course of events will still not reveal itself. There are no
governing social or economic laws to learn, and sensible agents will come
to recognize their capacity to make their own history in the context of contemporaneous institutions.
The fact that most macroeconomic time series are non-stationary provides empirical evidence for this view. Indeed, Solow (1985, p. 328) recognized as much when he wrote much of what we observe cannot be treated
as a stationary stochastic process without straining credulity. What is
more, while the existence of co-integration or unit roots may be suggestive
of an underlying ergodic relationship it may also be misleading, not least
for the introduction of spurious stationarity. As Klein (1994, p. 37) argues,
Stationarity means that in a time distribution of data, one could get the same
moments of the distribution no matter what block of time. It is a mathematical
property of a time series or other kind of collection of sample data. I do not
think economic data are necessarily stationary or that economic processes are
stationary. The technique of co-integration, to keep dierencing data until stationarity is obtained and then relate the stationary series, I think can do damage
. . . [as it] may introduce new relationships, some of which we do not want to
have in our analysis.
Nevertheless the concepts of stationarity and non-stationarity should not

be conflated with the ergodicnon-ergodic distinction. If the estimates of
the time averages do not vary with the period under observation then a stochastic process can be said to be stationary. However, as some stationary
stochastic processes are non-ergodic, that is, limit cycles, non-stationarity is
not a necessary condition for the existence for non-ergodic processes. But all
non-stationary processes are non-ergodic. Non-stationarity is thus a sucient condition for non-ergodicity and provides an empirical foundation for
Post Keynesian claims about the relevance of history and uncertainty.
Nevertheless Davidsons discussion seeks to go beneath purely stochastic
considerations, identifying the underlying causal mechanisms and emergent
properties and structures that generate non-ergodic time series. In a muchneglected aspect of his discussion of non-ergodicity, Davidson (198283, p.
192) emphasizes the link to G.L.S Shackles concept of creative, crucial decision making, arguing that the existence of crucial decision making represents
a sucient condition for the existence of non-ergodic processes. Situations
where purely processing information from the past provides insucient
information about the course of future events is suggestive of a creative role
282
Non-ergodicity
for human agency. Here crucial decisions refer to non-routine decisions that
take place in historical time. A crucial decision involves large transaction and
sunk costs and cannot be unmade without loss; it calls attention to the fact
that one is irrevocably tied (married) to ones decisions. Moreover, as this
conceptualization relates to both consumption and production decisions
that involve calendar time and large transaction costs, it moves beyond
Keyness apparently arbitrary distinction between autonomous investment
and non-autonomous consumption decisions. It provides a more appropriate framework within which to elaborate Keyness principle of eective
demand and to outline the relevance of liquidity considerations and their
nexus to the non-neutrality of money.
In linking non-ergodicity to Shackles concept of crucial decision making,
Davidson advocates a broader, creative view of agency than that contained
within mainstream models of human behaviour. Accordingly he has
expanded the concept to incorporate non-stochastic processes (Davidson
1991). Deterministic models of decision making which are elaborated in
logical time require Leonard Savages ordering axiom the presumption, at
least in principle, that agents can make a transitive ordering over all possible
outcomes. This involves a pre-programmed future and invokes a substantive
rationality that is inconsistent with the fecundity which Post Keynesians
impute a posteriori to agents. Post Keynesians recognize that it is impossible
to form a transitive ordering over a yet-to-be created future in which circumstances inconceivable at the point of origination emerge.
Subsequently, and in response to the numerous meanings that could be
imputed to non-ergodicity, as well as to encompass developments in complexity and chaos theory, Davidson (1996) has reformulated his discussion
in terms of a distinction between immutable and transmutable economic
processes. Immutability encompasses the ergodic and ordering axioms and
embodies the presumption of a programmed stable, conservative system
where the past, present and future reality are predetermined whether the
system is stochastic or not (ibid., pp. 48081). In immutable models,
history is predetermined and choice is neither genuine nor matters. Under
such a reformulation immutability refers to attempts to elaborate (real or
imagined) universal event regularities and to develop theoretical structures
of the general form whenever event (type) X then event (type) Y. Thus it
closely parallels Lawsons discussion of closed systems. In contrast, the
broader notion of transmutability encompasses the stochastic discussion of
non-ergodicity within a creative and emergent conceptualization of history
in which choice is genuine, matters, and can make a dierence in the long
run not least in aecting liquidity considerations and influencing the
employment path of an economy over time. On this view of economic processes, sensible agents recognize that the environment in which they make
Non-ergodicity 283
decisions is characterized by the absence of programmed and predetermined processes and is creative, open, emergent and uncertain.
This transmutable conception of economic processes provides for a delineation between the Post Keynesian approach to modelling and theorizing
about economic processes and that of many Austrians, New Classicals, New
Keynesian and New Institutionalist economists (see Table 1). It also underscores some of the methodological anities between Post Keynesianism
and the German historical school, the older institutionalists and critical
Table 1
A.
Conceptualizations of economic processes
IMMUTABLE REALITY (AN ERGODIC SYSTEM)
Type 1 In the short run, the future is known or at least knowable. Examples of
theories using this postulate are:
(a)
(b)
(c)
(d)
Classical perfect certainty models

Actuarial certainty equivalents, such as rational expectations models
New Classical models
Some New Keynesian theories
Type 2 In the short run, the future is not completely known due to some
limitation in human information processing and computing power. Examples of
theories using this postulate are:
(a)
(b)
(c)
(d)
(e)
(f)
Bounded rationality theory

Frank Knights theory of uncertainty
Savages expected utility theory
Some Austrian theories
Some New Keynesian models (e.g., coordination failure)
Chaos, sunspot and bubble theories
B TRANSMUTABLE OR CREATIVE REALITY (A NON-ERGODIC

SYSTEM)
Some aspects of the economic future will be created by human action today
and/or in the future. Examples of theories using this postulate are:
(a)
(b)
(c)
(d)
Keyness General Theory and Post Keynesian monetary theory

Post-1974 writings of Sir John Hicks
G.L.S. Shackles crucial experiment analysis
Old Institutionalist theory
Source: Reproduced from Davidson (1996, p. 485).
284
Non-ergodicity
realists, and provides for a renewed exchange of ideas with other potentially
compatible approaches (Dunn 2000).
This broader discussion of non-ergodicity underscores the Post
Keynesian view of economic time and its nexus to the macroeconomics of
modern credit-money production economies. It provides for a strong critique of the rational expectations hypothesis, not rejecting it on the basis
that it provides an unrealistic model of actual decision making, but rather
advancing a distinctive view of human agency that is broader than that contained within models of bounded rationality or complexity and provides a
challenge to the conventional wisdom that markets work best without
government intervention (Davidson 1996). Moreover it can be used in theorizing in a positive fashion to clarify the informational foundations of
monetary non-neutrality and transaction cost theory, as well as providing
for a strategic conceptualization of the modern corporation (see Dunn
2001 and the references contained therein).
As Sir John Hicks conceded in personal correspondence with Davidson
(12 February 1983): I have missed a chance, of labelling my own point of
view as non-ergodic. One needs a name like that to ram a point home. I had
tried to read a book on stochastic processes, but I was not sharp enough to
see the connections (italics added).
S P. D
See also:
Agency; Austrian School of Economics; Critical Realism; Expectations; Liquidity Preference;
Money; Time in Economic Theory; Uncertainty.
References
Davidson, P. (198283), Rational expectations, a fallacious foundation for studying crucial
decision-making processes, Journal of Post Keynesian Economics, 5 (2), 18297.
Davidson, P. (1991), Is probability theory relevant for uncertainty? A Post Keynesian perspective, Journal of Economic Perspectives, 5 (1), 12943.
(4), 479508.
Dunn, S.P. (2000), Wither Post Keynesianism?, Journal of Post Keynesian Economics, 22 (3),
34364.
Dunn, S.P. (2001), Uncertainty, strategic decision-making and the essence of the modern corporation: extending Cowling and Sugden, Manchester School, 69 (1), 3141.
Klein, L. (1994), Problems with modern economics, Atlantic Economic Journal, 22 (1), 348.
Parry, W. (1987), Ergodic theory, in J. Eatwell, M. Milgate and P. Newman (eds), The New
Palgrave: A Dictionary of Economics, Vol. 2, London: Macmillan, pp. 1848.
Solow, R.M. (1985), Economic history and economics, American Economic Review, 75 (2),
32831.
Pricing and Prices

Post Keynesians see the discipline of economics as being concerned with
explaining the process that provides the flow of goods and services required
by society to meet the needs of those who participate in its activities.
Consequently, Post Keynesian economic theory is the theoretical explanation of this social provisioning process in a capitalist economy. Hence Post
Keynesian theory is concerned with explaining those factors that are part
of the process, including the setting of prices by business enterprises.
The business enterprise is a specific social organization for coordinating
and carrying out activities associated with the provisioning process. It consists of an organizational component, a production and cost component, a
series of routines that transmit information (such as costs, sales and prices)
which enables workers and management to coordinate and carry out their
activities, and a management that makes strategic decisions about prices (as
well as investment). When making decisions, management is motivated by
dierent goals, such as growth of sales, developing new products, entering
new geographical regions or markets, generating dividends for shareholders, and/or attaining political power. Hence, management views price
setting, or pricing, as strategic decisions designed to meet these goals
(Eichner 1976; Lee 1998; Downward 1999).
To set a price of a product, the pricing administrators of the business
enterprise first determine its cost base. Utilizing costing procedures
derived from the management accounting procedures used by the enterprise, the pricing administrators determine the products average direct
costs (ADC), average overhead costs (AOC) and average total costs (ATC)
at normal output. The relevance of normal output is that it enables the
pricing administrators to determine the products normal costs. That is,
since ADC, AOC and ATC vary as output changes, it is necessary to select
a particular amount of output if costs for pricing are to be determined
before production takes place and the actual costs of production are
known. With the normal costs administratively determined, the pricing
administrators select a profit mark-up to be applied to the normal costs to
set the price. This pricing procedure means that the price of the good is
set before the good is produced and exchange takes place. The pricing
administrators then take the administratively-determined price (which is
determined outside the market) and administer it to (or impose it on) the
market.
At the centre of the pricing process are the administratively-determined
285
286
Pricing and prices
mark-up, normal cost and target rate of return pricing procedures. Markup pricing procedures consist of marking up average direct costs based on
normal output to set the price, with the mark-up being sucient to cover
overhead costs and produce a profit:
labour and material-based mark-up pricing: price(NADC)(1k) (1)
where NADC is normal average direct material and labour costs of the
product and k is the mark-up for overhead costs and profits.
Normal cost pricing procedures come in two forms. The most basic consists of marking up NADC to cover overhead costs, which gives normal
average total costs (NATC), and then applying the profit mark-up to
NATC to set the price:
normal cost pricing:
price[(NADC)(1g)](1r)
(2)
where g is the mark-up for overhead costs and r is the mark-up for profit.
A more detailed normal cost pricing procedure consists of applying the
profit mark-up to a completely delineated NATC to set the price:
normal cost pricing:
price(NATC)(1r).
(3)
Lastly, the target rate of return pricing procedure consists of marking up

NATC by a certain percentage to generate a volume of profits at normal
output that will produce a specific rate of return with respect to the value
of the enterprises capital assets connected with the production of the
product. That is, given the value of the capital assets (VCA) associated with
the production of the product, the pricing administrators want to obtain a
specific target rate of return (TRR) on those assets. Therefore, the amount
of profits required to meet the target rate of return is TRRVCAtarget
profits, Pt. To incorporate the target profit figure into the price, Pt is first
divided by normal output (no) to get the targeted costing margin, cmt, and
then divided by NATC to get the targeted profit mark-up (t):
target rate of return pricing: price(NATC)(1t)
target rate of return pricing: price(NATC)[1TRRVCA/(no)NATC].
(4)
Given the targeted profit mark-up, if the business enterprise produces at
normal output, enough profits will be generated to attain the desired target
rate of return on the capital assets (Eichner 1976; Lavoie 1992; Lee 1998;
Downward 1999).
Pricing and prices
287
Because actual output can dier from the normal output, the actual costs
of production can dier from normal costs, so that the actual profit or
target mark-up can dier from the profit or target mark-up used to set the
price. Consequently, in the context of the ebbs and flows of business activity and especially the business cycle, the business enterprise will not always
achieve its target rate of return or desired profits, sometimes being above it
and other times being below it.
The prices set by pricing administrators using normal cost-based
pricing procedures have five properties. The first property is that the price
is not based on or related to actual costs, and immediate or current market
forces do not aect the profit mark-up. That is, irrespective of the pricing
procedures used by pricing administrators, the shape of the enterprises
average direct cost curve or its average total cost curve is immaterial for
pricing purposes. The costs used for pricing are determined prior to production and are based on normal output. Consequently, the shape of the
ADC cost curve or ATC curve is not important for price-setting purposes.
Instead the price is based on normal costs, while actual costs vary around
it as actual output varies around normal output. As for the profit markup, the evidence strongly suggests that it remains stable for significant
periods of time so that in some cases it is considered customary by pricing
administrators; it is based on long-term competitive forces and will change
when those forces change; and it is unaected by momentary fluctuations
in sales. To explain theoretically the magnitude and the relative stability of
the profit mark-up, Post Keynesians have utilized either market structure
arguments or investment-determining mark-up arguments. However,
neither argument has much empirical support. Thus the profit mark-up
remains theoretically underexplained in Post Keynesian theory (Wood
1975; Eichner 1976; C apoglu 1991; Lavoie 1992; Sawyer 1995; Downward
1999).
Given normal costs and the stability of the profit mark-up, it follows that
the second property of administered prices is that they are stable in that
they remain unchanged for extended periods of time and for many sequential transactions. Consequently, administered prices are neither exchangespecific prices nor prices that reflect the impact of immediate variations in
sales. This implies that markets that have stable, normal cost-based prices
are not organized like auction markets or oriental bazaars where the
retailer engages in individual price negotiation for each transaction. Rather,
an enterprise that desires to enter these unorganized markets must first
announce a price for its product and then enter into direct buyerseller
interaction to obtain sales. Since buyerseller interactions take place both
simultaneously and through time, business enterprises find that stable
prices are cost-ecient in terms of selling costs, reduce the threat of price
288
Pricing and prices
wars, and facilitate the establishment of goodwill relationships with customers (Lee 1998; Downward 1999).
A third property of administered prices is that they are set largely
without reference to an inverse pricesales relationship and are not set to
achieve a specific amount of sales. In studies of price determination, business enterprises have stated that variations of their prices within practical
limits, given the prices of their competitors, produced virtually no change
in their sales and that variations in the market price, especially downward,
produced little if any changes in market sales in the short term. Moreover,
when the price change is significant enough to result in a significant change
in sales, the impact on profits has been negative enough to persuade enterprises not to try the experiment again. Consequently administered prices
are maintained for a variety of dierent outputs over time. The fourth
property of administered prices is that they change over time. The pricing
administrators of business enterprises maintain pricing periods of three
months to a year in which their administered prices remained unchanged;
and then, at the end of the period, they decide on whether to alter them.
The factors which are most important to the enterprises in this regard are
changes in labour and material costs, changes in the mark-up for profit and
changes in normal output. Factors prompting the enterprises to alter their
profit mark-ups include short- and long-term competitive pressures, the
stage that the product has reached in its life cycle, and the need for profit.
Moreover, since normal output is administratively determined, it is possible
for pricing administrators to alter it perversely over the business cycle,
resulting in the NATC increasing in the downturn and decreasing in the
upturn. If the mark-ups for profit remain constant, then the pricing administrators would be setting countercyclical or perverse prices. Consequently,
administered prices can change from one pricing period to the next in any
direction, irrespective of the state of the business cycle. However, evidence
does suggest that within short periods of time (such as two-year intervals),
changes in costs will dominate price changes, whereas over longer periods
of time changes in the mark-up will play a more important role (Lee 1998;
Downward 1999).
The fifth and final feature of administered prices is its role in the reproduction of the business enterprise. That is, pricing administrators use costbased pricing procedures to set prices that would enable the enterprise to
engage in sequential acts of production over time and thereby reproduce
itself and grow. More specifically, because market conditions facing the
enterprises many products are not uniform and change over time, its
pricing administrators utilize a variety of multi-temporal, open-ended
pricing strategies designed to achieve time-specific and temporally undefined goals. The compendium of pricing strategies is known as the enter-
Production
289
prises pricing policy, and the prices that the pricing administrator administers to the various markets, are based on one or more of these strategies.
Thus, the administered prices of a business enterprise are strategic prices
whose common and overriding goals are reproduction and growth (Eichner
1976; Lavioe 1992; Lee 1998; Downward 1999).
F S. L
See also:
Competition; Income Distribution; Kaleckian Economics; Production.
References
Capoglu, G. (1991), Prices, Profits and Financial Structures, Aldershot: Edward Elgar.
Downward, P. (1999), Pricing Theory in Post Keynesian Economics, Cheltenham, UK and
Eichner, A.S. (1976), The Megacorp and Oligopoly, Cambridge: Cambridge University Press.
Elgar.
Lee, F. (1998), Post Keynesian Price Theory, Cambridge: Cambridge University Press.
Sawyer, M.C. (1995), Unemployment, Imperfect Competition and Macroeconomics, Aldershot:
Edward Elgar.
Wood, A. (1975), A Theory of Profits, Cambridge: Cambridge University Press.
Production
It can be argued that, in the mainstream neoclassical or marginalist theoretical apparatus, the role of production is subsidiary to that of exchange.
The core of that approach, which explores how scarce commodities are
allocated among alternative uses through the price mechanism, can be
represented using a model of exchange with given endowments of nonproduced commodities. Production can be introduced into this model at a
later stage to show how resources are transformed into goods through the
production function, without fundamentally altering the basic insights to
be drawn from the approach. In contrast, in Post Keynesian economics, as
in the classical political economy and Marxian/radical approaches, production takes a more central role.
The centrality of production in Post Keynesian economics can in turn be
related to some of its main concepts, such as eective demand, historical
time and uncertainty (see Dutt and Amadeo 1990). Perhaps the key concept
common to all varieties of Post Keynesianism is that of eective demand,
and its role in determining employment and unemployment in the short run
and the rate of accumulation and growth in the long run. Since eective
demand determines these by determining the level and rate of change of
production, production naturally takes a central place in the Post
290
Production
Keynesian approach. Another important idea stressed in Post Keynesian

economics is the concept of historical time, as opposed to logical time in
which historical processes and irreversibilities are not adequately captured.
Chick (1983, pp. 1621) discusses how major economic decisions such as
consumption, saving, investment and especially production can be portrayed as being made over historical time. In this sequence, production decisions have to be made prior to sales, but with the expectation of sales
aecting how much firms produce. These expectations have been referred
to in the literature as short-period expectations, which are dierent from
the long-period expectations which govern investment decisions, following
Keyness Marshallian approach of the General Theory. A related concept,
stressed more in some varieties of Post Keynesianism than in others, is that
of uncertainty, which is distinguished from risk because objective probabilities cannot be assigned to the consequences of many kinds of economic
decisions as is assumed in the analysis of risk. While the concept of uncertainty is stressed in discussions of investment and asset-holding decisions,
it has also been invoked in discussion of production. In uncertain situations
firms and other economic agents are often seen as following conventions
and rules of thumb, behaviour which may be much more rational than
doing detailed costbenefit calculations of their actions.
An examination of the theory of production at the level of individual
producers illustrates the importance of some of these ideas. Since Post
Keynesian economics does not comprise a unified body of theory, but
rather several dierent approaches, it is useful to consider two dierent
approaches to such a theory, one derived from the Marshallian tradition of
Keyness General Theory presentation, and the other the Kaleckian
approach which stresses imperfect competition. Although hybrid forms
which combine features from these two approaches exist, it is instructive to
consider the two separately.
In the Keynesian approach the firm is assumed to operate in a purely
competitive environment, in the sense that it expects to sell any amount at
the going price. But since the firm has to make its production decision
without knowing the price which will prevail when its produce will be
brought to the market it is assumed to form short-period expectations,
which take the form of an expected price. With the money wage assumed
to be given, the firm is then taken to maximize its profit by equating its marginal product of labour (assuming diminishing returns and a given stock of
capital) to the ratio of the wage to the expected price. This determines the
firms market-period or day equilibrium level of employment and output.
Once each firm makes its production and employment plans and carries
them out, income flows are generated in the form of wages and profits, and
these determine the level of consumption, while firms make investment
Production
291
plans depending on their long-period expectations, which are taken to be

exogenous. Assuming that aggregate eective demand depends on the price
level (perhaps because of its eect on the real wage, and hence on consumption demand with dierential propensities to consume out of wage and
profit income), the price is assumed to vary to equate demand to total
market-period equilibrium. There is no guarantee that the expected price
of firms (for simplicity assumed to be the same for all firms) will be equal
to the market-period equilibrium price. If they are dierent, firms will
adapt by adjusting their expectations, and thereby (under certain conditions) arrive at the short-period equilibrium level of production, at which
not only does the market clear but the firms short-period expectations are
also fulfilled, although long-period expectations are still taken as given.
The short-period equilibrium level of output depends on the level of investment spending, among other things, and may well be below the fullemployment level of output. This approach can be seen as providing a
simple formalization of the role of eective demand in determining output,
of historical time in which dierent periods are carefully distinguished, and
of uncertainty, through its invocation of short- and long-period expectations. Moreover, the approach can be used as a basis for examining changes
in long-period expectations and its relation to short-period expectations,
which can be shown to lead to various kinds of path-dependencies in the
determination of the aggregate level of production (see Dutt 1997).
However, some Post Keynesians exhibit some hostility to it and to the
aggregate demandaggregate supply analysis related to it, given its closeness to marginalist traditions following from its assumptions of pure competition and production functions that exhibit diminishing returns.
These Post Keynesians prefer the Kaleckian approach, in which firms in
oligopolistic situations enjoy some degree of monopoly power. In this
approach, given the uncertainty concerning the level of aggregate demand
and the behaviour of other firms, firms use the rule of thumb of setting
their price as a mark-up on their unit prime or variable costs. The assumption of a fixed unit labour requirement, a fixed money wage, that labour is
the only variable factor, and a fixed mark-up (which depends on factors
such as the level of industrial concentration), the price becomes constant,
and the firms adjust their level of output to the level of demand for their
product, while maintaining excess capacity given their stock of capital. In
this approach the level of production for each firm is determined by the
demand for the firms product. The aggregate level of production in the
economy is therefore determined by aggregate eective demand, and will in
general be consistent with excess capacity and unemployed labour. This
approach represents a more radical departure from neoclassical economics,
because of its assumption of fixed unit labour requirements, which can be
292
Production
allowed to vary due to changes in technology and in the social relations of

production as in Marxian and radical presentations, rather than due to
factor substitution.
The feature common to both approaches is the role of aggregate eective
demand. There cannot be a self-contained microeconomic theory of production: the level of output depends on the demand for each firm, which
depends on macroeconomic factors (and some rule according to which
total demand is apportioned between firms). This essentially macroeconomic theory of output determination makes output depend on aggregate
levels of consumption, investment and other sources of demand.
However, the discussion presented so far leaves open the possibility that,
although unemployment equilibrium can occur in the short run, there may
be forces in the economy which change the level of production of firms in
a direction which will drive aggregate production to fully employ all the
economys resources in the longer run. For instance, this may occur in the
first model if unemployment leads to a fall in the money wage which will
induce firms to employ more workers and to produce more. In both models
unemployment can lead to a fall in the wage or the price level, which
increases the real supply of money, which through wealth eects on
demand or through a reduction of the interest rate can induce firms to
invest more, and thereby increase the demand for goods. These forces can
take the economy to positions of full employment, as suggested by the neoclassical synthesis or Bastard Keynesian approach. The demonstration
that there are no such necessary tendencies is therefore an important
concern of Post Keynesian economics.
Post Keynesians argue that the money wage is determined by institutional factors rather than by the automatic forces of supply and demand.
Thus, the importance given to relative wages by workers in the wage bargaining process may prevent wage reductions, as argued by Keynes (1936)
himself, or issues such as eciency wages or insideroutsider considerations may explain wage rigidity. While this wage rigidity interpretation of
Keynesian macroeoconomics is common in mainstream circles, especially
in neoclassical synthesis and New Keynesian quarters, this interpretation
would make Keynesian economics little dierent from that of the preKeynesians, who were quite aware that wage rigidity could result in unemployment.
Post Keynesians have also emphasized that wage reductions need not
take the economy to full employment, following Keyness own lead in
chapter 19 of the General Theory. Thus, they argue that money supply in a
credit-money economy is demand determined, so that a fall in the wage and
price level, rather than automatically reducing the interest rate because of
an excess supply of money, will simply reduce the supply of money endog-
Production
293
enously. Moreover, even if the interest rate does fall, when the wage and
price levels fall, firms caught in an uncertain situation may not increase
investment, and asset-holders may simply wish to hold more money. Thus,
standard mechanisms of expansion relying on asset market considerations
are short-circuited. A fall in wages and prices may actually reduce the cash
receipts of firms from the sale of goods and, given precommitted costs,
might lead them to cut back investment, and even worse, declare bankruptcy in extreme cases. More generally, deflation will redistribute wealth
from debtors to creditors, thereby possibly reducing aggregate demand. A
fall in the money wage, if it results in a fall in the real wage, can also redistribute income from wage-earners to profit recipients with a lower marginal
propensity to consume, which also reduces aggregate demand. Moreover,
neo-Ricardian Keynesians argue that a fall in the interest rate may not
increase aggregate investment if one takes into account the fact that capital
goods are produced inputs, and that changes in the interest rate or profit
rate can cause changes in the relative prices of capital goods and lead to
perverse changes in aggregate investment demand. In the absence of automatic tendencies which take output to full employment, it may be supposed
that this will be achieved by governments, especially through fiscal and
monetary policies. However, such policies can be ineective or slow to take
eect, and there may be political constraints on full-employment policies
as discussed by Kalecki (1971), which imply that even in the longer run,
production is determined by aggregate demand considerations rather than
supply-side factors.
A K D
See also:
Eective Demand; Employment; Expectations; Kaleckian Economics; Marginalism; Sraan
Economics; Time in Economic Theory; Uncertainty.
References
Chick, Victoria (1983), Macroeconomics After Keynes. A Reconsideration of the General
Theory, Cambridge, MA: MIT Press.
Dutt, Amitava Krishna (1997), Equilibrium, path dependence and hysteresis in postKeynesian models, in P. Arestis and M. Sawyer (eds), Markets, Unemployment and
Economic Policy: Essays in Honour of G.C. Harcourt, Vol. 2, London: Routledge,
pp. 23853.
Dutt, Amitava Krishna and Edward J. Amadeo (1990), Keyness Third Alternative? The NeoRicardian Keynesians and the Post Keynesians, Aldershot: Edward Elgar.
Kalecki, Michal- (1971), Selected Essays on the Dynamics of the Capitalist Economy,
Keynes, John Maynard (1936), The General Theory of Employment, Interest and Money,
London: Macmillan.
294
Prots
Profits
While profits are literally the dierence between the revenue from sales and
the costs of production, the scope of the term varies according to what is
admitted as a cost. If one believes that all value is produced by labour (aside
from natural non-reproducible resources), then only labour incomes are
costs and all non-labour incomes accrue as profits. If, on the other hand,
tangible capital is thought to contribute towards the production process,
then dividends and other returns to the owners of plant and machinery are
netted out of profits.
Profit, under any economic definition, represents a return from financing
acts which produce a good or service. It does not, in economics, include
winnings from zero-sum activities such as gambling or arbitrage. Nor
would most economists accept that it includes net capital gains, although
the dierence between current profits and capital gains is temporal (higher
profit expectations raise the current value of an asset) and may be irrelevant to the firm if the asset is realized. Profits are a flow arising from current
production. Lack of clarity among authors about what they include in the
term profits, as well as lack of recognition of the non-uniqueness of the
definition, can be a source of confusion for readers and remains a barrier
to inter-school-of-thought discussions.
Profits are received for advances of money or resources made in the
expectation of future benefits, whether these moneys are used to buy the
services of current or intermediate inputs into the production process, as
exemplified by David Ricardos corn model, or so-called investment goods
whose use may last beyond a single production cycle. However, a convention used often in economic analysis (but not accounting), is to regard
profits earned on working capital as secondary details, in order to focus on
the more complex relationship between investment goods and profits. This
convention was probably established because expenditures made with
respect to longer time horizons exhibit greater volatility and uncertainty.
Accordingly investments which are expected to furnish returns beyond a
year will, in general, be more important determinants of other economic
phenomena than investments that repay themselves within a week.
From a single firms perspective, funds advanced in each time period
equal the amount of working capital advanced to cover the costs of producing the good or service: payments to labour, leasing costs (direct or
implicit owner costs) of using reproducible plant, equipment or intangible
assets, and rents on non-reproducible inputs. From the point of view of the
whole economy, however, rents are not true costs but transfer payments (as
there are no opportunity costs), and the costs of producing the reproducible assets can be decomposed in a similar way according to the costs of
Prots
295
the respective investment-goods businesses. Taken to its limit, each production process can be reduced to labour and non-reproducible inputs, with the
only true cost of production being a dated series of labour inputs, as
revealed by Sraas (1960) scheme of prices. However, in each stage of production, a surplus exists after the firm has paid for the costs of labour and
non-labour inputs. And this, summed over all firms, is gross aggregate
profit.
If the costs (Alfred Marshalls quasi-rents) of using capital goods, that
is, capital consumption or depreciation, are essentially released flows of
stored labour, what then accounts for the existence and size of profits? Why
isnt all income paid to labour? In order to understand the determinants of
profits it is important to recognize the importance of the ownership of
financial capital and the power of finance over economic resources.
Rentiers and firms will not invest in production, or any given project, unless
they expect it to return a minimum or normal rate of profit (the ratio of
profits accrued per funds advanced). Rentiers and firms are not compelled
to invest and they can, when they desire, keep their finances as secure financial assets. Furthermore, economic resources cannot be mobilized without
financial backing. Borrowing can supplement collateral but cannot replace
this backing (see Kalecki 1939). Hence, profits form the incentive for rentiers to invest.
The two components of ex ante normal profits include default-free interest payments and returns to compensate for the normal uncertainty associated with doing business. The default-free interest rate is set by central
banks and represents a totally secure alternative way for capital owners to
place their funds. Owners of financial assets, or their banking intermediaries, will not lend to businesses for investment unless they are assured of a
greater return than what they would receive from these default-free bill
rates. To the extent that these central banks oer an elastic supply of these
bills, the rate of interest becomes the minimum supply price for rentiers of
financial capital. However, non-zero default-free banking rates constitute a
net injection of funds into the economy and the cost to the public of this
convention or policy action may be a redirection of funds from the central
governments normal budgetary policies. That is, the higher are the interest
payments governments need to make to meet their interest obligations, the
more funds are diverted, in the first instance, from their appropriations to
portfolios.
Even if the default-free rate was set at zero, businesses would still require
a positive return in order to commit themselves to production. There are
substantial uncertainties associated with investments, and unless some
compensation is forthcoming a rational rentier would place funds in a
sequence of government bills instead. While the premium for uncertainty
296
Prots
is undoubtedly positive (to the extent that business people are economically
rational), there are few theories to explain how much is required to compensate the rentier for his/her risk of investing in a business, as the evaluation of uncertainty is subjective. Knight (1921) recognized that, at the limit,
profits are not required to compensate for risk (actuarial-based risk), for if
people have complete information about the probabilities of all possible
contingencies, then they can objectively estimate an expected value. The
greater the frequency of repeated instances of the risky situation, the more
certain will be the expected outcome. Accordingly, at the limit, no premium
is required to compensate for actuarial risk, only non-actuarial uncertainty.
Clearly if the business expects a project to return greater than normal
profits it will proceed with the project and consider the windfall (pure)
above-normal profits as a reward for recognizing an overlooked and unexploited opportunity. In this way, expected profits attract the attention of
entrepreneurs who will shift resources into or out of markets with a speed
determined by the magnitude of the dierence between demand and supply.
However, the distinction between monopoly profit and normal returns
is not as clear as it sounds. Where does the normal premium for uncertainty end and above-normal profits begin? Some industries are more inherently risky than others, and firms which aggressively pursue monopoly
profits by investing in uncertain and unpredictable intangible capital,
would expect to be compensated by a higher premium for uncertainty. Until
this issue has been resolved, no explanation can exist for the level of normal
profits and it would be dicult in practice to identify whether firms or
industries are receiving monopoly profits.
While positive ex ante profits are required for ex ante investment, these
profits do not always materialize. Ex post profits can vary for reasons
related more to macroeconomic factors than the behaviour or expectations
of the specific firm. For example, as recognized by Kalecki, while firms can
set their profit mark-up on unit costs, they cannot determine how much
they will sell and consequently how much total profit they will make.
There are relatively few theories which seek to explain the size of ex post
aggregate profits and thus the extent to which ex ante expectations can be
simultaneously realized (long-run equilibrium). Neoclasssical aggregate
production function theories, such as Robert Solows 1956 model, have
been used in conjunction with J.B. Clarks marginal productivity theory to
show that the normal rate of profits is simply the value of the marginal
product of (aggregate) capital. As such, at long-run equilibrium the level of
profits reflects the innate productivity of aggregate tangible plant and
equipment. However, all measures of aggregate capital (needed to calculate
the value of the marginal product of capital) use the prices of capital goods
and thus an embedded rate of profit. This endogenous value measure of
Prots
297
aggregate capital does not therefore constitute an independent explanation

of the rate of profit.
The nineteenth-century theory of Ricardo defined the level of profits per
unit of output to be the dierence between the subsistence wage for labourers per unit of output and the average product of labour on the most marginal land in the economy. The level of production was consistent with full
employment. However, this theory is less relevant where the average wage
rate can and does vary over time and where subsistence is culturally defined.
Kalecki (1939) has one of the most comprehensive theories of the determination of profits. Using a simple two-sector macroeconomic model (with
no government or foreign sectors) he uses the two identities:
Y Cw CK I
YWP
where Y is output, Cw is workers consumption, CK is capitalists consumption, I is investment demand, W is the wage bill and P is total profits. If it
is assumed that workers do not save, such that CW W but capitalists
consume a small portion of their income, such that Ck AP (where A
is a given constant and is the marginal propensity to consume) this gives
P(AI)/(1). Since I can be determined by the deliberate decisions of
businesses (and A and by rentiers) but P cannot, the direction of causation must run from I, A and to P. Introducing workers saving obscures,
but does not destroy, this basic conclusion. Kalecki also had a microeconomic theory of the minimum ex ante rate of profit (which he took as exogenous to the economy) and average profit per unit of output (which
depended on the rate of competition). However, the ex post rate and level
of profits for each firm depend on the aggregation of all microeconomic
investment decisions, and accordingly are out of the hands of any individual business.
Finally, a word about how entrepreneurs or business managers actively
seek to make profits. Classical and neoclassical theories commonly portray
the flow of profit-seeking funds as action like water passively seeking an
even level. In contrast to this, contemporary Post Keynesian, Austrian and
evolutionary theories, which have a genesis in Joseph Schumpeter and
G.L.S. Shackle, endow the entrepreneur with a more aggressive and less
mechanical role. These theories begin with the postulate that any activity,
which aims to maximize profits, implicitly aims to maximize monopoly
profits. Monopoly or above-normal profits are received through exploiting
some special demand or cost advantage that creates economic distance
between the firm and its nearest market rivals. Thus a profit-seeking firm is
ultimately aiming to develop endogenous barriers to entry.
298
Prots
This assertive profit-seeking behaviour by firms involves intangible

investments in the development of marketing and distribution channels,
R&D, workforce training and management strategies (see Webster 1999).
Investing in tangible plant and equipment per se will not create abovenormal profits, as there is nothing unique or dicult to copy about such
activities and they do not therefore result in the creation of endogenous
barriers to entry. Tangible capital can be bought o the shelf and reproduced ad infinitum at a constant cost. Intangible assets, by contrast, are
heterogeneous and dicult to copy primarily because they are heavily
embodied in diverse human beings. It is investment in intangible capital
that creates monopoly profits and endogenous barriers to entry and thus
monopoly profits.
Essentially, this brings the discussion of the source of profits back to the
original point that profits are incomes accruing to non-labour inputs. If, on
the one hand, profits are received by owners of financial wealth but, on the
other, their size depends on the behaviour of other human beings, then an
incongruity may exist within the incentive structure between the creators of
monopoly profits and their beneficiaries.
E W
See also:
Capital Theory; Competition; Income Distribution; Innovation; Investment; Kaleckian
Economics; Sraan Economics; Uncertainty.
References
Kalecki, M. (1939), Essays in the Theory of Economic Fluctuations, Oxford: Clarendon Press.
Knight, F.H. (1921), Risk, Uncertainty and Profit, Boston: Houghton Miin.
Sraa, P. (1960), Production of Commodities by Means of Commodities, Cambridge:
Webster, E.M. (1999), The Economics of Intangible Investment, Cheltenham, UK and
Rate of Interest
Interest is the price for the use of capital the pure remuneration of
capital whatever the form of its employment, whether financial or real. If
production is carried on with the firms own capital, interest constitutes its
opportunity cost and as such will enter into that normal cost which in the
long run tends to be equated with the unit price. Firms would not continue
to replace plant which is wearing out unless the prices for their commodities were such that they could not do better for themselves by investing
their depreciation funds in gilt-edged securities; conversely, commodity
prices could not permanently involve rates of return on the firms funds
exceeding the relevant rates of interest those to be earned in the market
on long-term fixed-interest securities in which there is no element of risk
by more than a normal remuneration for the risk and trouble of productively employing capital. The case of share capital does not alter the fundamentals of this picture. It may be presumed that the nearest competing
alternative to shares is long-term bonds, and that ordinary shares will be
held only if the expected yield on them exceeds the yield on long-term
bonds. As there is a significant section of the investing public ready to
switch from one kind of investment to the other, this tends to maintain their
respective yields at a steady level. That is to say, at any given time there will
be a certain relationship between the prices of the various classes of securities: a shift in the price of one large class must be followed by a general
shift in the whole range of prices. Thus a rise in prices for long-term government bonds a fall in the long-term rate of interest resulting from the pursuing of a cheap-money policy will be followed by a rise in prices of
securities generally. But a higher quotation for existing equities implies that
companies can raise capital by issuing shares on more favourable terms; in
the words of Keynes, a high quotation for existing equities has the same
eect as if (companies) could borrow at a low rate of interest (1936, p. 151
n. 1). So the issue of common stock, as a method of financing investment
available to joint-stock companies, will also become cheaper (or dearer) in
the face of a persistent fall (or rise) in interest rates. We may conclude,
therefore, that quite irrespective of the kind of capital employed in production a lasting lowering (or raising) of interest rates tends to make normal
costs stand lower (or higher) than they would otherwise have done, and
thus, by the competition among firms within each industry, to aect prices
correspondingly. Given the level of money wages, any such change in the
price level brought about by a lasting change in interest rates would then be
299
300
Rate of interest
accompanied by a change in the same direction in the level of prices in relation to the level of money wages, thereby causing changes in income distribution. A prolonged fall in interest rates should cause a fall in prices relative
to the wage level and thereby bring about a lower rate of profit and a higher
real wage, while a prolonged rise in interest rates should raise the rate of
profits, and thus reduce the real wage.
Although economic theory has always looked at interest as the price for
the use of capital in production, it has however also generally regarded it
as a subordinate phenomenon. In the words of Joan Robinson: [o]ver the
long run, the interest rate rentiers can exact is dominated by the profits that
entrepreneurs can earn, not the other way around (1979, p. xxii). In fact,
according to both classical and marginalist economists there is, between the
normal rate of profit and the money rate of interest, a long-run causal relationship going from the former to the latter, so that the rate of interest is
ultimately determined by those real forces which explain the course of the
normal rate of profit: the real wage rate and production techniques, in the
classical theory of distribution up to David Ricardo; the fundamental phenomena of productivity and thrift, as far as marginalism is concerned. An
important implication of this way of conceiving the relation between interest and profit is the denial of any substantial power on the part of the monetary authorities. Given the state of the real forces governing normal profit
the natural real rate the impact on the price level or on real output and
accumulation of any lasting discrepancy between the courses of the two
rates would force the monetary authorities to act so as to make the rate of
interest move in sympathy with the rate of profit. An autonomous lowering of the lending rate by the monetary authorities would drive the price
level up, contrary to what has been outlined in the previous paragraph; this
is because overall monetary expenditure would expand as a consequence of
the dierence which would be created between the lending rate and the
natural real rate. In actual experience, however, rising prices very rarely
coincide with low or falling interest rates and the opposite is the general
rule (the so-called Gibson Paradox). Instead of assuming a lowering of
interest rates by the monetary authorities, other things being equal, one
would then simply have to make the alternative assumption that a dierence between the rate of profit and the rate of interest generally arises
because it is the former which rises or falls, while the latter remains
unchanged and only belatedly follows (see Wicksell 1906 [1962], pp. 2045).
An unprejudiced observation of concrete reality clearly played a significant part in Keyness interpretation of the rate of interest as a monetary
phenomenon. The fact that Keynes was far from being entirely happy with
his monetary explanation of interest (Keynes 1937 [1973], p. 213), did not
shake his conviction that the rate of interest is not determined by the real
Rate of interest
301
forces envisaged by the neoclassical theory. Unfortunately, the persistence

in Keyness analysis of some traditional neoclassical premises seriously
weakens his concept of the rate of interest as a magnitude determined by
monetary factors. In particular, the idea of an investment demand schedule constitutes an obstacle which a monetary theory of interest cannot
easily overcome. Notwithstanding the statement in the General Theory that
he no longer regards Wicksells concept of a natural real rate as a most
promising idea (Keynes 1936, p. 243), the natural rate is still there, as the
rate that would ensure equality between full-employment saving and investment decisions. Keyness underemployment equilibrium is ultimately the
result of the presence in the economic system of factors that hinder the possibility of bringing the actual rate of interest down to its natural or fullemployment level. It is, in other words, the result of a limited flexibility of
the money rate of interest. This limited flexibility is actually all that Keynes
has to oer as a basis for his non-orthodox concept of interest as a monetary phenomenon. But if one takes into account the fact that even in
Wicksell there is no automatic gravitation of the money rate towards the
level of the natural real rate (banking policy having to perform the task),
then the dierence between the two authors will not appear that marked.
They both share the idea of an inverse relation between the rate of interest
and investment decisions, while the conflict of opinions is essentially
centred upon the degree of the (non-automatic) flexibility of the rate of
interest, in the face of discrepancies between full-employment savings and
investment decisions. One can say that it was largely in the light of this comparison that the neoclassical synthesis could argue, successfully and with
foundation, that the determination of the current rate of interest by the
intersection of the supply schedule of money and the demand schedule for
money, while adequate for showing that the flexibility of the rate of interest is not of an automatic nature, is, however, insucient to sustain the
thesis of a limited flexibility of the rate of interest. And if current money
interest can normally be brought to, and kept at, its natural level provided the monetary authority applies to its action a modest measure of
persistence and consistency of purpose (Keynes 1936, p. 204) then the
neoclassical real forces of productivity and thrift may still be regarded as
the ultimate determinants of the equilibrium rate of interest.
Things are quite dierent if there is no such thing as a natural rate of
interest a normal rate of profit, that is to say, determined independently
by real forces and which can be taken as the primum movens. We would be
back in this case to the picture outlined at the beginning of this entry, that
is, it would be dicult not to acknowledge that, given money wages and
production techniques, a lowering (raising) of interest rates by the monetary authorities would actually drive the price level down (up), owing to the
302
Rate of interest
adaptation of prices to normal costs caused by competition. There would

thus be nothing paradoxical in the positive correlation between interest
and prices one generally finds in actual experience. And at a given level of
real output, the rate of interest would also regulate the quantity of money
in active circulation a quantity that adapts itself to the needs of trade
via its influence on the price level: interest, prices and the quantity of
money would all move in the same direction, with the policy-determined
interest rate acting as the primum movens of the process.
The monetary explanation of distribution that I started to develop a
few years ago (see Pivetti 1991) is precisely an attempt to emancipate us
from any real explanation of the rate of interest an attempt prompted by
Sraas suggestion that in the necessary long-run connection between
normal profit and money interest it is the latter which is susceptible to
setting the pace (see Sraa 1960, p. 33). By focusing on the actual mechanism whereby the rate of interest is likely to set the pace in its connection
with normal profit, eventually I got hold of the notion of money interest
as an autonomous determinant of normal money production costs which
governs the ratio of prices to money wages. As pointed out earlier, this
interpretation of interest does not require any particular assumption as to
the kind of capital employed in production: borrowed, in the form of
shares or a firms own capital. For any given situation of technique, there
is a price level that depends on the money wage and on the money rate of
interest, with the latter acting as the regulator of the ratio of the price level
to the money wage. This ratio is thus seen as the connecting link between
the rate of interest and the rate of profit: by the competition among firms
within each industry, a persistent change in the rate of interest causes a
change in the same direction in the level of prices in relation to the level of
money wages, thereby generating a corresponding change in the rate of
profits and an inverse change in the real wage. Wage bargaining and monetary policy come out of this analysis as the main channels through which
class relations act in determining distribution. Class relations are seen as
tending to act primarily upon the profit rate, via the money rate of interest, rather than upon the real wage as maintained by both the English classical economists and Marx. Indeed, the level of real wage prevailing in any
given situation is viewed as the final result of the whole process by which
distribution of income between workers and capitalists is actually derived.
Interest rate determination is thus not seen as constrained by a normal
profitability of capital which is predetermined by some natural, technical
or accidental circumstances be they the relative scarcity of capital and
labour, a subsistence real wage, or the rate of growth of the economic
system. Rather the rate of interest is regarded as a policy-determined variable, a conventional monetary phenomenon determined from outside the
Rate of interest
303
system of production (Sraa 1960, p. 33) and not subject to any general
law. One can describe interest rate determination in terms of sets of objectives and constraints, on the action of the monetary authorities, which have
dierent weights both among the various countries and for a particular
country at dierent times (see Pivetti 1991, pp. 1117, 336), and with
which, to a very large extent, the parties relative strength is ultimately
intertwined.
M P
See also:
Bastard Keynesianism; Endogenous Money; Income Distribution; Liquidity Preference;
Monetary Policy; Sraan Economics.
References
Macmillan, 1964.
Keynes, J.M. (1937), Alternative theories of the rate of interest, Economic Journal, 47 (186),
24152. Reprinted in The Collected Writings of John Maynard Keynes. Volume XIV: The
General Theory and After. Defence and Development, D. Moggridge (ed.), London:
Macmillan and Cambridge University Press for the Royal Economic Society, 1973,
pp. 20115.
Pivetti, M. (1991), An Essay on Money and Distribution, London: Macmillan.
Robinson, J. (1979), The Generalisation of the General Theory and Other Essays, London:
Macmillan.
Wicksell, K. (1906 [1962]), Lectures on Political Economy. Volume II: Money and Credit,
London: Routledge & Kegan Paul.
Saving
What is the relationship between saving behaviour in capitalist economies
and their macroeconomic performance? This question is a hardy perennial
in the history of economics, and one that has carried great theoretical and
practical significance through its many revivals. It is easy to see why this is
so, since any economy that aspires to long-term increases in productivity
and average living standards must devise eective means of raising the
quantity and quality of its capital stock. The role of saving is central to this
process, though how exactly it exerts influence has long been a matter of
contention.
Debates on how saving behaviour aects long-term growth and business
cycles stretch back to those between David Ricardo and Thomas Malthus
on whether Says Law of markets that supply creates its own demand can
be violated, thereby creating the possibility for general gluts or depressions. The Keynesian revolution, of course, was also focused on this issue,
as Keynes rebelled against the 1934 Treasury View that higher saving rates
were a necessary precondition for stimulating investment and lifting the
British economy out of depression. Arguing against the intuitively appealing notion that an adequate pool of saving must exist before the funds for
investment can be drawn, Keynes and Richard Kahn developed the concept
of the multiplier to demonstrate the counterintuitive point that higher
levels of investment will generate higher saving as well. Many of the most
pressing policy concerns of today remain centred on the relationship
between saving and macroeconomic performance.
What is saving? The answer is not obvious. Moreover, answering the most
basic questions about the impact of saving on macroeconomic activity
including whether saving rates are rising or falling depends on how one
defines and measures the term (this discussion follows Pollin 1997b). Two
standard approaches to measuring saving are as an increase in net worth
and as the residual of income after consumption. As accounting categories,
these two saving measures should be equal in value. But making this distinction raises a major question: when one considers the category of assetspecific saving, should the value of assets be measured at historical costs or
market values? Only the historical cost measure is equivalent conceptually
to residual saving. Measured at market values, asset-specific saving will of
course fluctuate along with fluctuations in asset prices themselves.
Another major issue is distinguishing gross saving, including depreciation allowances, and net saving, which excludes depreciation. In principle,
304
Saving
305
net saving measures the funds available to finance economic growth, while
gross saving would also include funds set aside for replacing worn-out
capital stock. In practice, however, depreciation allowances do not simply
finance replacement. Rather, they are primarily used to finance investment
in capital stock that represents some advance over previous vintages. As
such, depreciation funds are also utilized to promote economic growth.
What is the most appropriate definition of saving? In fact, for the purposes of economic analysis, there are legitimate reasons to examine each
concept. There are three basic reasons for considering saving patterns by
any measure. The first is to observe households portfolio choices, in which
case asset-specific saving is obviously the only option. The second purpose
would be to understand consumer behaviour. Here we would want to
measure consumption directly relative to income, making saving a residual.
However, asset-specific saving at market values would also be important
here in so far as it contributes to understanding consumer behaviour. The
third reason for measuring saving is with respect to examining its role in
determining credit supply, that is, the source of funds available to finance
investment and other uses of funds. This role for saving is clearly the
primary consideration among analysts seeking to understand the relationship between saving and macroeconomic performance. In fact, however, the
connections between any given measures of saving, the provision of credit,
and overall rates of economic activity are quite loose. We can see some indication of such loose connections through Table 2, on the US economy.
Table 2 Saving rates, credit supply and GDP growth for the US economy
(in percentages)
Net private saving/GDP

Gross private saving/GDP
Net worth private saving/GDP
Total lending/gross private saving
Real GDP growth
196069
197079
198090
19912000
9.6
17.1
25.2
60.5
4.4
9.8
18.4
35.2
86.9
3.3
8.9
19.1
32.4
106.1
2.9
6.3
16.4
30.5
106.4
3.2
Note: For brevity, two sets of cycles (197073/197479 and 198081/198290) have been
merged.
Sources:
US National Income and Product Accounts; US Flow of Funds Accounts.
The first three rows of the table show annual figures on net, gross and net
worth saving in the US relative to nominal GDP between 1960 and 2000.
The data are grouped on a peak-to-peak basis according to National
Bureau of Economic Analysis business cycles. The last two rows of the
306
Saving
table show, respectively, measures of credit supply and overall activity: first
the ratio of total lending in the US economy relative to gross private saving,
then the average annual growth rate of real GDP.
To begin with, the table shows substantial dierences in the cycle-tocycle behaviour of the three saving ratios. For example, between the 1970s
and 1980s cycles, the net saving ratio fell from 9.8 to 8.9 per cent, the gross
saving ratio rose from 18.4 to 19.1 per cent, and the net worth ratio fell from
39.5 to 32.4 per cent. Meanwhile, between these same two cycles, the
lending/saving ratio rose sharply from 86.9 to 106.1 per cent, while the rate
of GDP growth fell from 3.3 to 2.9 per cent.
At the very least, one can conclude from these patterns that we cannot
take for granted any analytic foundation through which we assume a simple
one-way pre-Keynesian causal connection whereby, as James Meade (1975,
p. 82) put it, a dog called saving wagged its tail labelled investment instead
of the Keynesian connection in which a dog called investment wagged its
tail labelled saving.
The pre-Keynesian orthodox view held that the saving rate is the fundamental determinant of the rate of capital accumulation, because the saving
rate determines the interest rate at which funds will be advanced to finance
investment. Keyness challenge to this position constituted the core of the
ensuing Keynesian revolution in economic theory. Nevertheless, what we
may call the causal saving view was nevertheless restored fairly quickly to
its central role in the mainstream macroeconomic literature.
Despite neglect among mainstream economists, the causal investment
perspective has advanced substantially since the publication of Keyness
General Theory (1936). One major development has been precisely to establish a fuller understanding of the interrelationship among saving, financial
structures and real activity. This has brought recognition that the logic of
the causal investment position rests on the analysis of the financial system
as well as the real-sector multiplieraccelerator model.
Of course, the multiplieraccelerator analysis is the basis for the
paradox of thrift, that is, low saving rates (saving as a proportion of
income) can yield high levels of saving and vice versa when real resources
are not fully employed. However, considered by itself, the multiplieraccelerator analysis neglects a crucial prior consideration: that the initial increment of autonomous investment must be financed, and the rate at which
financing is available will influence the size of this investment increment
and the subsequent expansion of output, income and saving.
Kaldor (1960) was an early critic of the multiplieraccelerator causal
investment position, and his argument was revived by Asimakopulos
(1983). Their critique focuses on the interregnum between an autonomous
investment increase and the attainment, through the multiplieraccelerator
Saving
307
process, of a new level of savinginvestment equilibrium. The Kaldor

Asimakopulos position is that, as a general case during such interregnum
periods, intermediaries could not be expected to accept a reduction in
liquidity without receiving an interest rate inducement to do so. Rather, for
intermediaries to supply an increased demand for credit would require
either a rise in interest rates or a prior increase in saving. As such, low rates
of saving again yield high interest rates and a dampening of investment
an argument, in other words, that returns us to the causal saving position.
In fact, Keynes himself addressed this issue, working from his theory of
liquidity preference and interest rate determination. But this dimension of
his argument is far less well known than the consumption function and
multiplier analysis, at least in part because it is less fully developed in the
General Theory itself.
Holding the level of saving constant, Keynes argued that the banking
system private institutions as well as the central bank was capable of
financing investment growth during the interregnum without necessarily
inducing a rise in interest rates. That is, as he put it, In general, the banks
hold the key position in the transition from a lower to a higher scale of
activity (1973, p. 222). Keynes based his position on a central institutional
fact, that private banks and other intermediaries, not ultimate savers, are
responsible for channelling the supply of credit to non-financial investors.
The central bank can also substantially encourage credit growth by increasing the supply of reserves to the private banking system, thereby raising the
banks liquidity. But, even without central bank initiative, the private intermediaries could still increase their lending if they were willing to accept a
temporary decline in their own liquidity. The reason that the fall in the
intermediaries liquidity would be only temporary is that liquidity would
rise again, even before the completion of the multiplier, when the recipients
of the autonomous investment funds deposited those funds with an intermediary. Moreover, the completion of the multiplier process would mean
that an increase in saving equal to the investment increment had been generated. Overall, then, it is through this chain of reasoning that Keynes
reached what he called the most fundamental of my conclusions within
this field, that the investment market can become congested through a
shortage of cash. It can never become congested through a shortage of
saving (1973, p. 222).
This more fully developed Keynesian position emphasizes clearly the
central role of financial institutions in establishing the relationship between
saving and macroeconomic activity. More recent literature has developed
this idea in several directions (see the contributions in Pollin 1997a). Other
researchers have broadened further this investigation as to the relationship
among saving, institutional structures and macro activity. Indeed, in the
308
Saving
1990s a substantial literature developed arguing that financial systems that

channelled savings within a tighter regulatory structure tended to outperform economies in which capital markets operated more freely (Pollin 1995
reviews this literature). Countries categorized as having more tightly regulated bank-based financial systems were Germany, France, Japan and,
among the less-developed Asian countries, South Korea. The US and the
UK represented the less-regulated capital marked-based system. But, by
the mid-1990s, the debate over the relative merits of these systems was
short-circuited by two factors: first the stock market bubble in the United
States, which lent temporary credence to the idea that capital market-based
systems could operate more eectively; and, second, the global ascendance
of neoliberal economic policies in economies such as Japan, France and
Korea, contributing, in turn, to greater economic instability in these economies in the late 1990s. But a restoration of this line of research on alternative financial institutional environments will be critical for developing
new policy regimes that can promote more stable as well as more egalitarian growth prospects.
More broadly within the realm of policy, there always have been clear
important normative issues at play in the debates over saving behaviour.
The agenda following from a causal saving perspective consists of seeking
to raise national saving rates through measures such as providing preferential tax treatment to capital income, eliminating government deficit spending, or even paying o completely outstanding government debts. These
will normally also generate a less equal distribution of income. Building
from a causal investment analytic framework points to policy approaches
that directly encourage higher investment while also promoting egalitarian
distributional outcomes. Such measures would include increasing aggregate demand and employment through fiscal and monetary interventions
or more equal income redistribution, or, through various institutional
reforms, giving preferential access to credit for productive private investment relative to unproductive speculative expenditures. The policy ideas
that flow from a causal investment perspective are committed to utilizing
most eectively the interconnections observed in research among growth,
stability and distributional equity.
R P
See also:
Consumption; Keyness General Theory; Multiplier; Says Law.
References
Asimakopulos, A. (1983), Kalecki and Keynes on finance, investment and saving, Cambridge
Journal of Economics, 7 (34), 22133.
Says law
309
Kaldor, Nicholas (1960 [1939]), Speculation and economic activity, in N. Kaldor, Essays on
Economic Stability and Growth, London: Duckworth, pp. 1758.
Keynes, John Maynard (1936), The General Theory of Employment, Interest and Money, New
Keynes, John Maynard (1973 [1938]), The ex ante theory of the interest rate, in The
Collected Writings of John Maynard Keynes, Volume XIV: The General Theory and After:
Defence and Development, London: Macmillan for the Royal Economic Society, pp. 21523.
Meade, James (1975), The Keynesian revolution, in M. Keynes (ed.), Essays on John
Maynard Keynes, Cambridge: Cambridge University Press, pp. 828.
Pollin, Robert (1995), Financial structure and egalitarian economic policy, New Left Review,
214, NovemberDecember, 2661.
Pollin, Robert (ed.) (1997a), The Macroeconomics of Saving, Finance and Investment, Ann
Arbor: University of Michigan Press.
Pollin, Robert (1997b), Financial intermediation and the variability of the saving constraint,
in Pollin (ed.), pp. 30966.
Says Law
Says Law, also known as the law of markets, is a set of ideas and propositions that were originally formulated during the classical period in the
history of economics. Despite the name, Jean-Baptiste Say was neither the
inventor of the law nor was he its clearest and most coherent advocate.
Important contributions to the development of the law were made by Adam
Smith, James Mill, David Ricardo and John Stuart Mill. After the demise of
classical political economy and the rise to dominance of neoclassical economics, Says Law remained essentially unchallenged, even though economists paid much less attention to it. Says Law has always generated
opponents. During the classical period, Thomas Robert Malthus, J.C.L.
Simonde de Sismondi and Karl Marx were some of the most important
critics; in modern times, John Maynard Keynes undoubtedly provided the
most radical and clearly articulated critique of Says Law. Only Keynes,
however, succeeded in convincing a significant part of the profession that the
law was incorrect and had to be rejected (for a thorough historical reconstruction of the debate on Says Law, see Sowell 1972, but also Baumol 1977).
The basic idea underlying the law of markets is that there cannot be any
obstacle to economic growth deriving from an insucient level of aggregate demand. Whatever the level of aggregate supply, it will give rise to an
equal level of aggregate demand. If there are obstacles to growth, they
depend on other factors, such as, for Ricardo, decreasing returns in agriculture. In the analyses of markets carried out by classical adherents to Says
Law, the production and sale of goods generates an income which is either
spent for consumption or saved. What is saved, however, is also spent, as it
is devoted to investment. Thus production gives rise to purchasing power
of equal value that is entirely spent to buy the current production itself.
310
Says law
Situations in which aggregate supply exceeds aggregate demand are therefore impossible.
The analysis was based on the assumption that the exchange of goods
through money is conceptually the same as barter, in which it is impossible
that any divergence between demand and supply arises because sellers are
necessarily at the same time buyers. Money was regarded merely as a device
to facilitate exchanges, which was therefore demanded only for this function. It was assumed that people do not draw any utility from holding
money in a larger amount than that required to exchange commodities. The
law in its classical formulation also implied that the existing productive
capacity of the economy is always fully utilized, even though this does not
mean that labour is fully employed: the existing productive capacity might
be insucient to employ the entire labour force. In other words, if there is
unemployment, it does not result from an insucient level of demand but
from an insucient growth of capital.
It is hard to find Says Law expressed in these terms by any post-classical
economist. None the less, neoclassical economics left the law of markets
essentially unchallenged. Its basic aspects were accepted: money was still
regarded essentially as a mere device to make exchanges more ecient, and
it was held that saving is necessarily transformed into investment. However,
there were some significant dierences between the classical and neoclassical versions of the law. First of all, classical economists simply assumed
that saving is investment, so that any discrepancy between the two variables
is impossible; neoclassical economists, instead, admitted the possibility
that saving and investment diverge. Such a divergence, however, would be
eliminated by some equilibrating mechanism. Second, for the classics, the
law only implied the full employment of capital; whereas, for neoclassical
economics, the validity of the law also implied the full employment of
labour. Also in this case, unemployment of labour would be eliminated by
an equilibrating mechanism.
For Smith, Ricardo and so on, the transformation of saving into investment was essentially a direct process, in the sense that savers themselves
were those who invested. For the neoclassicals, the process was essentially
indirect: saving and investment decisions are not necessarily made simultaneously and by the same people. For this reason saving and investment may
diverge; but market mechanisms take care of this by ensuring that the
equality is restored. The variable that plays this equilibrating role is the rate
of interest. Abstaining from present consumption (saving, that is, the
supply of capital) is a direct function of the interest rate, and investment
(the demand for capital) is an inverse function, so that any divergence
between the two variables is eliminated by variations in the interest rate. For
the classics, the possible existence of unemployment did not give rise to any
Says law
311
adjusting mechanism that would bring the economy to the full employment
of labour; for neoclassical economists, variations of the (real) wage rate
bring the economy to full employment. If there is an excess supply of
labour, a decrease in the real wage rate would induce firms to increase their
demand for labour and eliminate unemployment.
These analytical dierences between classical and neoclassical economists can also be pointed out by expressing Says Law in two dierent ways:
as an identity and as an equality. Classical economists, by assuming that
saving is investment, accepted the law as an identity (the equality between
aggregate supply and demand is always true); neoclassical economists, by
concentrating on equilibrating mechanisms, accepted the law as an equality,
which is true only in equilibrium (see, for example, Sowell 1972, pp. 348).
Keynes held that, before him, Malthus went closest to a satisfactory criticism of Says Law, but he was probably far too generous to Malthuss critique of Say and Ricardo. Malthus did not provide a coherent alternative
analytical framework, as he maintained the assumption that saving is
investment. It was Marx who developed a more satisfactory critique of the
law of markets, which resembles Keyness own criticism in several respects
(Sardoni 1991). In Marx, as well as in Keynes, the rejection of Says Law is
based on the idea that the analysis of the working of the economy must be
carried out by taking account of money, which plays a more crucial role
than merely being a device to facilitate exchange. For Marx, the existence
of money breaks the unity between selling and buying; the exchange of
commodities through money is not conceptually the same as barter. Those
who own money can always decide not to convert it into commodities and,
hence, break the unity of exchange. In particular, capitalists can decide to
keep money idle instead of investing, whenever they expect that producing commodities will not be profitable.
Turning now to consider Keynes, it is useful to distinguish between two
dierent analytical levels at which he developed his critique of Says Law.
On the one hand, he carried out a critique of the law that is addressed to
the essential theoretical foundations on which it rests; on the other hand,
especially in The General Theory, he also developed a criticism that is more
complex and articulated, as it takes into account analytical aspects that are
typical of the neoclassical version of the law.
Keyness essential critique hinges on the idea that what makes the law
untenable is that the economy of the world in which we live is a monetary
economy, that is, an economy in which money plays a much more fundamental role than the advocates of the law implied. The income of the
factors of production, which is generated by production, is not necessarily
spent entirely on current output; income not spent on consumption goods
can be kept in the form of money rather than being transformed into goods.
312
Says law
The explanation of why individuals may wish to hold money as a store of

value is one of the key elements of Keyness theory. He was deeply convinced that the economic and social environment is dominated by uncertainty, which cannot be reduced to risk and treated with the traditional
tools of probability theory. But, notwithstanding uncertainty, individuals
have to make decisions and to act. They do so by pretending that the calculation of the probabilities of a series of prospective advantages and disadvantages is possible. In order to behave in such a way, some techniques
are devised, which essentially are conventions like assuming that the
present is a reliable guide to the future despite the past evidence to the contrary, or trying to conform to the behaviour of the majority.
These are, for Keynes, flimsy foundations for decision making, and they
are subject to sudden and violent changes. It is in this context that money
plays a crucial role, dierent from its function as a medium of exchange.
Money demanded as a store of value is, for Keynes, an indicator of peoples
distrust of their conventions concerning the future. When conventions
break down and expectations become more uncertain, demanding money
is a sort of insurance against uncertainty. The demand for money,
however, is in no way similar to the demand for any other good or service:
an increase in the demand for it, which means a decrease in the demand for
some other goods or services, does not give rise to a corresponding increase
in its production. Money has a zero elasticity of production (Keynes 1936,
p. 230). Thus, as in Marxs analysis, money can be demanded and kept idle
rather than being spent on goods or services. This is the basic reason why
Says Law does not hold.
Keynes, however, criticized the law in its neoclassical version; therefore,
he also had to consider those specific aspects on which neoclassical economists concentrated. In particular, he had to reject the idea that there is an
equilibrating mechanism that brings aggregate supply and demand to
equality by ensuring the equality between saving and investment. For
Keynes, saving is not necessarily transformed into investment. The act of
saving does not imply the supply of a corresponding amount of funds to
those who wish to invest. Once the amount of saving has been decided, the
individual has to decide whether to keep it in the form of money or to part
with it for a certain time, that is, to lend it. Such a decision depends on the
individuals liquidity preference. This vision of saving decisions implies that
the interest rate is not a return to saving on waiting as such. In fact, if a
man hoards his saving in cash, he earns no interest though he saves as much
as before. The interest rate, instead, is the reward for parting with liquidity (Keynes 1936, p. 167). Therefore, the rate of interest cannot play the
equilibrating role that is given to it in neoclassical analysis. There can be situations in which liquidity preference is so high that the interest rate is at too
Socialism
313
high a level to allow investment to reach its full-employment level, that is,
that level that ensures the level of aggregate demand associated with the full
employment of labour and capacity. In other words, it is not true that any
level of supply whatsoever generates an equal level of demand.
Keynesian economists of the neoclassical synthesis accepted Keyness
rejection of Says Law but in their analyses, based on the ISLM model, they
concentrated on the more technical aspects of the reasons why the interest
rate cannot guarantee the full-employment level of investment (for example,
the so-called liquidity trap), while Keyness deeper critique of market economies and his notion of uncertainty were essentially neglected. Post
Keynesian economists, by emphasizing the importance of decision making
under uncertainty, more clearly link the rejection of Says Law to an alternative notion of the essential features of market economies. Contemporary
mainstream economics, having rejected Keyness theory altogether, has
returned to the full acceptance of Says Law, even though macro economists
hardly ever mention it in their analyses.
C S
See also:
Bastard Keynesianism; Keyness General Theory; Liquidity Preference; Rate of Interest;
Uncertainty; Unemployment; Walrasian Economics.
References
Baumol, W.J. (1977), Says (at least) eight laws, or what Say and James Mill may really have
meant, Economica, 44 (174), 14562.
Macmillan.
Sardoni, C. (1991), Marx and Keynes: the critique of Says Law, in G.A. Caravale (ed.), Marx
and Modern Economic Analysis, Vol. II, Aldershot: Edward Elgar, pp. 21939.
Sowell, T. (1972), Says Law. An Historical Analysis, Princeton: Princeton University Press.
Socialism
Most Post Keynesians are not socialists. But almost all are reformers, who
advocate major reforms of capitalism, designed not only to help stability, but
also to increase equality. There is nothing incompatible between advocating
major reforms in the short run and advocating a complete change to socialism in the long run. So a minority of Post Keynesians are also socialists.
What is meant by socialism? Fortunately, the Soviet Union is dead and
buried. So we can ask what a Post Keynesian approach to socialism may
be. If one wants to consider something called socialism, there are three
questions to be asked: the role of democracy in socialism, the plan versus
the market in socialism, and the degree of equality in socialism.
314
Socialism
All Post Keynesian socialists are agreed that any socialist society must be
democratic. To understand the issue fully, however, the long and convoluted history of this issue must be briefly mentioned. Karl Marx and
Friedrich Engels fought in the 1848 revolution for democracy in Germany
and they supported every movement for democracy in their lifetime.
European socialist parties, from the German Social Democrats to the
British Labour party grew up in a struggle for the extension of democracy,
including surage rights for male workers and for women.
After the Bolshevik revolution of 1917 in Russia, however, the embattled
Bolsheviks became less and less democratic, with 60 years of one-party dictatorship. Their defence was a parody of Marxism: they argued that the
Soviet Union had only one class, the working class, so it needed only one
party to be democratic. In truth, it was a dictatorship over workers and
everyone else and the horrors of that dictatorship set back the cause of
anything called socialism by many decades.
Socialism means that the people rule over the economy, not a small group
of capitalists or a small group of bureaucrats appointed by a dictator. A
precise definition would be that socialism means the extension of democracy from the political sphere into the economic sphere. Some of the procedural necessities and safeguards in the political sphere under socialism
are spelled out in detail by Ralph Miliband (1994). In the economic sphere,
two types of democratic procedures have been advocated. One is to have
local, state and national ownership by a democratically elected government. The other procedure is to have corporate boards democratically
elected by the employees of each enterprise. Of course, there may be any
mix of these two forms even a role for a democratic government representative and regulations within an employee-run enterprise. This poses the
question of the manner of control of the economy.
Post Keynesian socialists are divided on what would be the best way to
coordinate a socialist economy. Some probably a minority at present
favour some form of democratic central planning. Others perhaps a
majority favour a market type of socialism (for a good introduction to
the debate among socialists, see Ollman 1998).
The Soviet Union had extreme central planning with dictatorship. In that
model, almost all industry and about half of agriculture was directly owned
and run by the government. A group of planners, appointed by the ruling
party, drew up a plan for the whole economy. To do this, they had to know
(i) the available resources, including all known raw materials, every type and
quality of labour, and every type and quality of capital. They also had to
know (ii) the preferences of the government and all consumers. Finally,
they had to know (iii) all the technological coecients telling them what
could be produced with a given amount of resources. Of course, even with
Socialism
315
the best presently available information network, collecting all such information down to the enterprise level would be impossible in any limited
amount of time, such as six months to a year. So they dealt instead with
broad aggregates of each industry, divided at most into about 500 categories. With that general information, they calculated several variants of a
national plan. It was then up to the government dictatorship to decide on
which variant would be used, depending on its time preferences between
present production and investment for future expansion. The plan which
was adopted was then handed down as law to the managers appointed by
the dictatorship. Although some discussion was allowed in the early stages,
criticism of the final plan was not allowed, nor was it healthy for employees to criticize the managers.
Such excessive planning in a dictatorial setting nevertheless managed to
transform the economy from an underdeveloped, mainly rural and agricultural one, to the second largest industrialized economy in the world. From
1928 to 1989 excluding the Second World War the Soviet economy never
suered a decline in aggregate output. Moreover, it had full employment
that entire time. So it has been argued that, whatever its other problems,
central planning can provide for full employment, development out of
underdevelopment, and rapid growth rates for at least a time. On the other
side, it was clear that the Soviet economy suered from a high level of ineciency. That ineciency grew worse as the economy became more complex
and the problems of central planning grew. So it has been argued that
central planning always leads to enormous ineciency though whether
it is worse than the crises of capitalism is another question. The truth,
however, is that the Soviet experience gave us very limited lessons. It was
not democratic planning, but planning under a one-party dictatorship.
Dictatorships must lead to vast ineciency in government-owned enterprises. There can be no criticism of cabinet decisions on the plan, there can
be no criticisms of the details of planning at the lower levels, and there can
be no criticism of management (until after they are fired). There can be no
freedom for scientists to choose the direction of scientific research because
the ultimate decisions on the direction and funding of research will be
under political control. Thus one cannot say whether Soviet ineciency
resulted from excessive planning, from dictatorship, or from their combination. If it was the combination, how should the blame and the praise be
allocated?
Those who argue that central planning always leads to unacceptable
ineciency argue that socialism would be better with a market form.
Managers could be told to maximize profit, rather than follow a plan. But
most advocates of market socialism would argue for employee-controlled
enterprises, since this would provide a high level of direct democracy. Of
316
Socialism
course, government could control some functions, such as environmental

regulations, while employees controlled the rest. The only extensive experience known of a form of market socialism with employee control of enterprises was in Yugoslavia but that again was conditioned by a one-party
dictatorship that aected the results. Yugoslavia did seem to have considerable enterprise eciency, such as better quality control than the Soviet
Union. But as Yugoslavia moved from central planning to independent
enterprises, it also witnessed all the problems of capitalism, including
monopoly profits, inflation, and cycles of boom and bust with cyclical
unemployment. It also moved towards a society with great emphasis on
consumerism, and with feelings of helplessness by unemployed or employees in firms doing poorly or going bankrupt. There was plenty of food and
other goods in the stores, but an unequal income distribution which meant
that many people could not buy an adequate basket of goods. Many people
began to work in more than one job. So market socialism has been attacked
as leading to instability and alienation.
Many Post Keynesian socialists conclude that, under present technological and social circumstances, a feasible socialism must combine market
and plan. The smallest businesses could remain private. Medium-sized
businesses could all be employee controlled so as to provide incentives and
eciency and internal democracy. To stabilize the economy, the giant conglomerates require direct public control and planning. Even the largest,
government-owned firms would leave many functions, such as safety, hours
and forms of the production process, to internal democracy. This is not a
vision of utopia, but it would mean the end of extraction of profits from
employees by a small elite and it would mean a large degree of economic
democracy through planning by democratic governments at local, state,
national and world levels, as well as a large measure of internal democracy
for firms with the most employees.
There is vast inequality under capitalism. John Maynard Keynes advocated greater equality of income distribution, both to increase the stability
of capitalism and for ethical reasons. All Post Keynesians advocate more
equality under capitalism and would surely endorse more equality under
socialism.
Socialism by itself does not necessarily mean equality. If all capitalist
ownership and profits are eliminated, then the single largest source of
inequality will be eliminated. In addition, democratic control of the
economy may bring a decision for greater equality, but it could mean a decision to retain wide disparities of income. Under centrally planned socialism, where all income is determined by planners and politicians, it is
possible for a democratically elected government to reduce drastically the
inequality of wages and salaries, but it is also possible to increase wage and
Socialism
317
salary inequality. In a market socialist economy, the market decides; so one

person might earn $5,000 a year, while another earns $1,000,000 a month,
In the old Soviet Union, Stalin decided that the conservative argument,
that great income inequalities are necessary to provide great incentives, was
correct. Therefore, for much of Soviet history, the degree of inequality of
wage and salary income was about the same as in the United States. It is
true that private profit income was outlawed, but the elite did receive secret
income and very important non-monetary privileges, including villas,
chaueured cars and medical specialists.
At the other extreme, in the utopia described by Edward Bellamy (1887)
there was central planning with equal wages for everyone with only a few
medals or fewer hours for the most dirty, dangerous and dicult jobs. But,
given the present psychology of employees under capitalism, it is true that
equal wages would enormously reduce incentives to work. Perhaps decades
of new experience and propaganda would change that psychology, but that
is speculation.
Still more utopian is the world described by Ursula Le Guin (1974), in
which there is central planning, but there is no money or prices, so everyone can take what they need with only peer pressure as a goad to work
or a limit on demand. With present psychology, that would mean almost
no work and unlimited demand an impossible situation. In addition,
aside from incentive and demand problems, central planning with no
explicit prices or money would be very dicult in any complex economy.
Most Post Keynesians would urge fiscal means to reduce inequality. They
advocate highly progressive taxes together with government spending to
provide certain free goods and services to everyone with negligible eect on
incentives. If the items are necessities, such as health care, then there will be
only a limited increase in demand.
Health care is provided to some extent by every industrialized country
other than the United States. In the United States, over 40 million people
have no guaranteed health care. If a socialist society provided free health
care to all, that would reduce inequality, provide for a basic need and
increase productivity. Another area of struggle in many capitalist countries
is the provision of free higher education. If a socialist society wished to go
further, it could ensure that everyone has a minimum necessary amount of
food to eat, a certain minimum level of housing, and free public transportation, with no economic disruption if this is done gradually.
H J. S
See also:
Kaleckian Economics; Transition Economies.
318
Sraan economics
References
Bellamy, Edward (1887 [1987]), Looking Backward, New York: Modern Library.
Le Guin, Ursula (1974), The Dispossessed, New York: Harper Prism.
Miliband, Ralph (1994), Socialism, New York: Verso.
Ollman, Bertell (1998), Market Socialism: The Debate among Socialists, New York and
London: Routledge.
Sraan Economics
Among the many unsettled issues within Post Keynesian economics, one of
the most divisive concerns the relationship between Keyness theoretical
framework and the analytical tradition associated with Piero Sraa
(18981983). In the 1970s, Sraan economics was generally regarded not
only as compatible with Post Keynesianism, but as an important branch of
it. (The essays in Nell 1980, for example were the culmination of a project
to integrate Keynesian, Kaleckian, Sraan and Marxian insights into a
unified Post Keynesian account of postwar capitalism.) By the end of the
1980s, however, this view had largely given way to the presumption that the
two frameworks are distinct and, in at least some key respects, incompatible.
Sraa is best known as the author of Production of Commodities by
Means of Commodities (1960), which provided the basis for a capital-theoretic critique of the neoclassical theory of distribution. Orthodox theory
explains income distribution in terms of the interaction of the demand for
and supply of scarce factors of production. The theory requires (i) that the
endowment of each factor of production be known prior to the determination of prices and distribution; and (ii) that the demand for any factor
declines as the price of its productive services increases. Sraas analysis
suggested that in a long-period setting neither of these conditions can be
presumed to hold for the factor called capital. Since capital is composed
of many dierent kinds of produced means of production, whose prices
themselves depend upon the real wage and the profit rate, the endowment
of capital, specified as a value magnitude, cannot be known prior to distribution. Moreover, changes in distribution can cause the prices of capital
goods to vary in highly complex ways, so that no systematic relation can be
established between the profit rate and the capital intensity of production.
The upshot of all this is that the substitution mechanisms that underpin the
price-elastic factor demand functions of neoclassical theory lack solid
foundations. But these substitution mechanisms are precisely what justify
the orthodox claim that, in the absence of impediments to the adjustment
of prices, market forces will push the economy towards full employment.
The capital critique was part of a larger constructive agenda that motivated Sraas scientific work from the late 1920s. His objective was to lay a
Sraan economics
319
foundation for the reconstruction of the classical political economy

approach pioneered by Adam Smith, David Ricardo and Karl Marx that
had been, as Sraa (1960, p. v) put it, submerged and forgotten with the
rise of neoclassical economics. A distinctive feature of the classical theory
is its treatment of income distribution not in terms of the equilibrating
interaction of price-elastic supply and demand functions, but as the
outcome of the interplay of class interests in a historically conditioned
institutional setting. Profits are conceived as a residual, or surplus, appropriated by the owners of capital after the replacement of the material inputs
including the wage goods consumed by workers used up in the production of aggregate output; for this reason the classical theory is sometimes
called the surplus approach.
In their analysis of value and distribution the classicals treated the following variables as parametric: (i) the size and composition of the social
product; (ii) the technical conditions of production; and (iii) the real wages
of workers. The equation systems of Parts I and II of Production of
Commodities establish that these data are sucient to determine relative
prices and the profit rate. (Sraas formulation fixes the profit rate and
determines the real wage as a residual; but it makes no dierence to the
logic of the theory which distribution variable is taken as parametric.)
From the same equations Sraa derived a trade-o between the real wage
and the profit rate, which corresponds to the classicalMarxian conception
of distribution as grounded in the opposition of class interests.
Sraas analysis and Keyness theory of eective demand intersect at
several junctures. First, by knocking out the foundations of neoclassical
distribution theory the capital critique buttresses Keyness contention that
there is no mechanism within capitalism capable of ensuring that the labour
market will tend to clear. Second, the classical theory of value and distribution including Sraas modern formulation of it is open-ended with
regard to the determination of the level and composition of the social
product. Some variation on the KeynesKalecki eective demand mechanism could therefore provide an explanation of outputs that is compatible
with the classical surplus framework.
Third, Sraas analysis calls into question some aspects of Keyness
articulation of the theory of eective demand, and helps to account for the
theorys assimilation, in weakened form, into mainstream economics.
Keynes incorporated into The General Theory two elements of Marshallian
orthodoxy: the proposition that in equilibrium the real wage must equal the
marginal product of labour; and the notion that investment demand is a
decreasing function of the interest rate. The capital critique undermines
both of these claims. The Sraan literature on Keynes maintains, furthermore, that by adopting them he opened the way for the interpretation of
320
Sraan economics
The General Theory in terms of the neoclassical synthesis, according to

which, under conditions of wage and price flexibility, the Keynes eect and
the real balance eect ensure that the labour market tends to clear in the
long run. On this interpretation Keyness argument applies to the short
period or to circumstances in which persistent market imperfections
prevent the price adjustments that would otherwise bring about full
employment. In failing to detach himself fully from his Marshallian roots,
Keynes produced an inherently unstable compromise that in the end fell
short of its revolutionary promise (Garegnani 197879; see also Milgate
1982).
Sraa appears to have harboured reservations about The General
Theory. His manuscripts at Trinity College, Cambridge, indicate that he
disliked the liquidity preference theory of interest, but this by itself seems
inadequate to account for his scepticism. It is possible that Sraa recognized earlier and more clearly than other members of Keyness circle that
the books argument was susceptible to assimilation into the orthodox
framework Keynes wanted to scuttle. The way past this diculty, most
Sraan writers would contend, lies in jettisoning the Trojan horses Keynes
incorporated into The General Theory the Marshallian elements that are
anyway vulnerable to the capital critique.
Throughout most of the formative period of Post Keynesian economics,
from 1960 until the mid-1980s, the anities between the KeynesKalecki
eective demand mechanism and Sraas work were not a matter of
dispute. Since then, however, the Post Keynesian project of integrating
Keynes, Michal- Kalecki and Sraa has lost momentum, and something of
a rift appears to have developed between the Sraan camp and many of
those who identify themselves as Post Keynesians. The history of this rift
remains to be sorted out, though it appears to be connected, no doubt as
both cause and eect, to the failure of the Trieste summer school to achieve
one its principal goals the forging of a unified methodological and
analytical foundation for the Post Keynesian project. The tensions are
largely unnecessary, and at least some Post Keynesian resistance to the
Sraan view is based on a misunderstanding of it. There are of course
genuine dierences of perspective, but these do not render the two traditions incompatible with each other.
Much of the Post Keynesian literature exhibits a strong antipathy
towards attempts to explain economic phenomena in terms of equilibrating forces, and the classical surplus approach has come in for some criticism because of its utilization of models of long-period gravitation to
explain aspects of income distribution and price determination. There is no
disputing the practical importance of the themes emphasized by Post
Keynesians uncertainty, expectations, disequilibrium; what is at issue is
Sraan economics
321
how to incorporate these phenomena into a coherent account of social processes. Uncertainty, technical change and coordination failure account for
much of the complexity of economic life. A Sraan would argue that it is
the very complexity of these issues that justifies the indispensability of a
long-period theory of value and distribution: disequilibrium processes can
best be understood by reference to the gravitational forces they disrupt.
There is no necessary incompatibility on methodological grounds between
the Post Keynesian and Sraan frameworks; the two complement each
other, each providing a dierent set of tools for dierent sorts of analytical problems.
Similarly, the Sraan approach to money is not incompatible with a
broad Post Keynesian perspective. The Sraan framework allows for the
non-neutrality of money via the eect of the money rate of interest on distribution. Pivetti (1985) and Panico (1985), developing a hint from Sraa
(1960, p. 33), have argued that the rate of interest fixed by the monetary
authorities regulates the rate of return on non-financial capital. The argument starts from the familiar stylized fact that central banks can regulate
interest rates but cannot generally control the money supply; the reasoning
is therefore consistent with the endogeneity of money, a defining Post
Keynesian premise. In the models developed along this line, money is not
neutral at all: the monetary authorities, by setting the interest rate, influence distribution, and this has consequences for the composition, level and
growth of aggregate output, and for employment. The forces that regulate
distribution are no doubt more complicated than the mechanism described
in these models, but the approach nevertheless clarifies important aspects
of the distributional role of finance. The question is not whether money is
non-neutral on this, the Sraan and Post Keynesian positions are not in
conflict but how it is non-neutral.
Underlying the Sraan argument is a presumption that what we can say
about money depends on how we conceive value and distribution. One of
the reasons the Post Keynesian literature relies so heavily on the pervasiveness of uncertainty is that it is not solidly grounded in a theory of value and
distribution; this deficiency is partly a byproduct of the rift with the
Sraans. Post Keynesian arguments about uncertainty are not so much
wrong as extremely limited in what they can explain. Long before the publication of The General Theory, Keyness neoclassical predecessors understood that uncertainty and fluctuations in business confidence could
disrupt the markets coordinating mechanisms. If, as many Post Keynesians
contend, the monetary nature of capitalist production is what accounts for
unemployment, it would appear to follow that in a non-monetary economy
the postulates of neoclassical theory hold, and that such an economy would
tend towards full employment. But the capital critique undermines the
322
Stagation
neoclassical theory of distribution, and hence its theory of employment,

whether the economy utilizes money or not. The Sraan analysis of money
and interest suggests furthermore that the non-neutrality of money does
not reside or at any rate does not reside exclusively in the uncertainty
inherent in monetary exchange regimes. Non-neutrality is indeed incompatible with the neoclassical theory of distribution; but there is no evident
incompatibility between the non-neutrality of money and the classical
surplus approach of Sraa.
Fuller discussions of these issues can be found in the essays contained in
Eatwell and Milgate (1983). For a general overview of Sraas scientific
legacy, see Mongiovi (2002).
G M
See also:
Cambridge Economic Tradition; Capital Theory; Equilibrium and Non-equilibrium; Income
Distribution; Rate of Interest; Uncertainty.
References
Eatwell, J. and M. Milgate (eds) (1983), Keyness Economics and the Theory of Value and
Distribution, Oxford: Duckworth.
Garegnani, P. (197879), Notes on consumption, investment and eective demand, Parts I &
II, Cambridge Journal of Economics, 2 (4) and 3 (1), 33553 and 6382.
Milgate, M. (1982), Capital and Employment, London: Academic Press.
Mongiovi, G. (2002), Classics and moderns: Sraas legacy in economics, Metroeconomica,
53 (3), 22341.
Nell, E.J. (ed.) (1980), Growth, Profits and Property, Cambridge: Cambridge University Press.
Panico, C. (1985), Market forces and the relation between the rates of interest and profit,
Contributions to Political Economy, 4, 3760.
Pivetti, M. (1985), On the monetary explanation of distribution, Political Economy: Studies
in the Surplus Approach, 1, 73103.
Stagflation
The term stagflation describes the simultaneous occurrence of inflation
and stagnation, the latter defined by the existence of high unemployment.
The term gained popularity in the early 1970s with the simultaneous
increase in rates of inflation and unemployment throughout the developed
capitalist economies. Hitherto in the post-Second World War period, the
behaviour of inflation and unemployment rates traced out a negative relation between inflation and unemployment rates, a relation formalized in the
Phillips curve, which was to become an integral part of Post Keynesian
macroeconomics.
Stagation
323
Two related impacts of the events following the Great Inflation of the
early 1970s were a serious challenge to the dominance of Keynesian macroeconomics and the emergence of a radically dierent set of beliefs.
Supposedly the simultaneous occurrence of rising unemployment and
inflation rates constituted definite proof of fundamental flaws in Keynesian
theory and its application as a tool for forecasting (Lucas and Sargent
1978). To Post Keynesians, and indeed to macro economists mindful of the
complexity of the economic world, this response was but another example
of the dangers of relying on gross correlations to substantiate or refute a
theory. In fact, by simply taking account of additional influences on inflation rates that were prominent during the stagflation period, the negative
relationship assumed in the original Phillips curve remains unchallenged.
In isolating the critical historical events responsible for stagflation, it is necessary to give a more precise definition of stagflation. Here the term will
refer to the simultaneous occurrence of high unemployment rates and high
rates of inflation compared to those of the golden age.
Figure 15 shows average annual rates of inflation p and unemployment U
of the seven large OECD (Organization for Economic Cooperation and
Development) economies from the closing years of the golden age to the
closing years of the last century. An outstanding feature throughout the
period in the G7 (and in other OECD economies) is the pronounced upward
trend in unemployment rates beginning in the late 1960s until 1983. A second
notable feature is the upward trend in inflation rates from 1967 until 1980,
covering most of the same period in which unemployment rates doubled
their golden-age rates and resulting in over a decade of stagflation.
Furthermore, during the first half of the 1980s unemployment rates rose to
triple their golden-age rates, and not until the mid-1980s did inflation rates
fall to their golden-age rates. Finally, Figure 15 reveals that, from the mid1980s until the recession of the early 1990s, the small decline in unemployment rates from historic highs was sucient to double inflation rates by 1990.
Only in the 1990s, with the return of unemployment rates to approximately
212 times their golden-age rates, did inflation fall back to its golden-age rates.
Events of the times raised several questions whose answers are still a
matter of dispute. First, why were unemployment rates not only high but
also trending upwards since the early 1970s? Second, if, as Post Keynesians
maintain, stimulative aggregate demand (AD) policies are eective in
reducing unemployment when unemployment is involuntary, why were
such policies not implemented? Third, what were the causes of a more than
decade-long period of stagflation? And, fourth, why did it require a period
of high unemployment of over three decades, culminating in almost two
decades of unemployment rates nearly 212 times their golden-age levels, to
rid the system of high inflation?
Percent
324
Stagation
13
12
11
10
9
8
7
6
5
4
3
2
1
U
p
p
1967 1970
1975
1980
1985
1990
1995
Note: Data include West Germany prior to 1993, after which data for the unified Germany
are used.
Sources: OECD, Historical Statistics, various issues; OECD, Economic Outlook, 44,
December 1988, Table R11; OECD, Economic Outlook, 65, June 1999, Annex Tables 16 and
22.
Figure 15 Average annual rates of inflation (p) and standardized

unemployment (U) for the G7 countries, 19671997
To those of a neoclassical persuasion, capitalism is a self-regulating
system. The core of their explanation of the long-run upward trend in unemployment rates was a rising equilibrium rate of unemployment, alternatively referred to as the NAIRU (non-accelerating inflation rate of
unemployment) or natural rate of unemployment. This is a rate of unemployment determined solely by supply forces, which AD policies can alter
permanently only by accepting ever accelerating or decelerating rates of
inflation. Hence no other rate of unemployment was natural. To Post
Keynesians, much if not most of the pronounced upward trend in unemployment since the early 1970s could be explained in terms of deficient aggregate demand. A sustained policy-induced increase in aggregate demand
could have led to a permanent reduction in unemployment without an acceleration of inflation rates, because the unemployment was largely involuntary. In other words the long-run Phillips curve was downward sloping. Why,
then, did the authorities allow the unemployment rate to remain so high?
Stagation
325
The opinion of governments and central bankers was that if stimulative

policies strong enough to maintain the economy at golden-age unemployment rates had been introduced any time before the late 1990s, inflation
would have risen to politically unacceptable, although not necessarily accelerating, rates. Unlike the golden-age period, when the goals of full employment and acceptable rates of inflation could be realized simultaneously, the
early 1970s saw the emergence of an inflationary bias, that is, an inability
of the economy to reach full-employment rates of unemployment without
experiencing unacceptable rates of inflation. A shift in the Phillips curve
outward to the right had taken place. Recovery in unemployment required
more than a stimulation of aggregate demand; it required measures to
contain inflation at full employment. When these proved unavailable, the
full-employment goal was sacrificed in the interests of restraining inflation.
This leads to the third question what initiated the episode of stagflation?
Figure 15 reveals a sharp rise in inflation in the late 1960s to rates well
above their previous golden-age rates, followed by a smaller decline in the
early 1970s and then, in 197374, the first of two explosions in inflation
rates. Underlying these price movements was a series of shocks, beginning
with the wage explosion of the second half of the 1960s, continuing with
the breakdown of the Bretton Woods agreement, the rapid run-up of commodity prices and prices of foodstus in the early 1970s and finally the
explosion in oil prices in 197374. In each case the impacts of the shocks
were then magnified by the activation of wagewage and wageprice inflationary mechanisms and propagated throughout the economies. By themselves the disturbances provide a plausible explanation of the Great
Inflation up to the mid-1970s and, when account is taken of a second oil
shock in 197980, of the strong inflationary trend until 1980. Thus, even
without taking further account of the wage explosion, the episode of stagflation could be explained as large price disturbances dominating the eects
of policy-induced unemployment increases on inflation outcomes.
Without denying the influence of shocks on inflation, such an interpretation ignores the lasting impact of developments leading up to the explosion
in money wages that had preceded the other shocks just cited. While the
increase in unemployment rates worked to restrain money and thereby price
increases, influences other than shocks and rising unemployment rates had
a major impact on wage settlements and therefore on price movements
throughout the period covered in Figure 15. These influences were trends in
underlying income and employment expectations and aspirations of the
average citizen and worker, generated by the prolonged period of low unemployment and rising living standards of the golden age. During the period
leading up to the Great Inflation, living standards had risen steadily and
were expected to continue to rise indefinitely. Furthermore, a growing pro-
326
Stagation
portion of the labour force had never experienced long spells of unemployment, if any, and were convinced that full employment would be guaranteed
by government. These growing expectations and aspirations generated
growing demands upon the system that would carry over beyond the golden
age. Thus the so-called wage explosion of the late 1960s was a symptom of
a profound institutional change that had a lasting eect on inflation and
unemployment.
Evidence supporting the lasting eects of rising income and employment
expectations and aspirations can be seen in the behaviour of wage and price
inflation in the mid-1970s. In 197576, wage and price inflation rates fell in
response to restrictive policies and a decline in commodity prices, including oil. However, only a slight reduction in unemployment rates in 197778
led to an increase in inflation rates once again. Events of the 1980s provided
further evidence of the short-lived eects of earlier restrictive policies on
wage and price inflation. Restrictive policies successfully reduced inflation
in the first half of the 1980s to levels little dierent from those of the golden
age, but at a large unemployment cost. However, beginning in 1987 with a
rather moderate fall in unemployment rates from existing historically high
rates, inflation rates rose until the recession of 1990.
This leads to the fourth question. Since the overriding concern of the
authorities over the period covered in Figure 15 was to reduce inflation
through restricting AD policies, eventually permitting a reduction in unemployment (short-term pain for long-term gain), why did the reduction of
inflation rates to their golden-age levels take so long and why was it so
costly in terms of unemployment? The mainstream explanation stressed the
lasting eects of the inflationary psychology that had built up over the
many earlier years of high inflation. Reversing the deeply ingrained expectations of continued high inflation would require an equally long period of
time and persistent restrictive policies.
An alternative explanation adhered to by Post Keynesians (Cornwall and
Cornwall 2001) and others (Phelps Brown 1971, 1975 and Salvati 1983)
focused on developments in the labour market, most of which have just
been discussed. As outlined, the transitory eect of restrictive policies and
the underlying strength of the average workers expectations and aspirations are related. In addition, the restrictive policies initiated to fight the
Great Inflation confirmed labours suspicion that they were made the
victims of the fight against inflation. This prolonged the period during
which labour would continue to demand high wage increases.
However, the impact of prolonged high unemployment and the stagnation of real earnings on income and job expectations and aspirations eventually worked to reduce wage demands. The high unemployment cost of
reducing inflation rates in the pre-1990s period can be attributed to the job
Stagation
327
and income expectations and aspirations formed in the golden age, and
continued into the earlier phases of the high-unemployment episode. The
ability eventually to bring down and to maintain low rates of inflation in
the 1990s can be attributed to an eventual reversal of expectations and aspirations. By the recession of the early 1990s, after 13 years of unemployment
rates averaging more than 212 times their golden-age rates, inflation rates
fell below their golden-age levels and remained so throughout the balance
of the decade. Unemployment rates fell in the second half of the 1990s, for
example in the United States, but on average the recovery in employment
in the G7 (and in the OECD) was modest. Stagflation had ended as inflation had been conquered but stagnation continued through the remaining
years of the 1990s (not shown in Figure 15).
No-one can predict the future but, barring the occurrence of unpredictable serious shocks or endogenous structural changes in the economy, there
seems no reason to expect these trends to be reversed in the near future. The
income and employment expectations and aspirations of the average
citizen are likely to remain subdued.
J C
See also:
Bretton Woods; Economic Policy; Inflation; Unemployment; Wages and Labour Markets.
References
Cornwall, J. and W. Cornwall (2001), Capitalist Development in the Twentieth Century; An
EvolutionaryKeynesian Analysis, Cambridge: Cambridge University Press.
Lucas, R. and T. Sargent (1978), After the Phillips curve: persistence of high inflation and
high unemployment, in After Keynesian Economics, Federal Reserve Bank of Boston,
pp. 4972.
Phelps Brown, E.H. (1971), The analysis of wage movements under full employment,
Scottish Journal of Political Economy, 18 (3), 23343.
Phelps Brown, E.H. (1975), A non-monetarist view of the pay explosion, Three Banks
Review, 105, 324.
Salvati, M. (1983), Political business cycles and long waves in industrial relations: notes on
Kalecki and Phelps Brown, in C. Freeman (ed.), Long Waves in the World Economy,
London: Butterworths, pp. 20224.
Taxation
It is dicult to find an explanation for the failure until recently of Post
Keynesian economics to develop its own approach to taxation. Afer all,
Kalecki had realized as early as 1937 that the publication of Keyness
General Theory required a whole new approach to taxation. His response
was a short paper in the Economic Journal entitled A theory of commodity,
income and capital taxation (Kalecki 1937). But Kalecki never developed
his thinking beyond the simple short-period model of his 1937 paper. In correspondence between Keynes and Kalecki at the time, Keynes indicated that
he was also thinking along similar lines to Kalecki, but had not developed
his ideas as far as Kaleckis. While a possible explanation for Keynes was
that he never completely broke free from the strictures of neoclassical
theory, particularly income distribution theory, this clearly does not explain
Kaleckis failure. The only significant postwar advance in the development
of a Post Keynesian approach has been by A. (Tom) Asimakopulos and J.
Burbidge (1974). However, theirs was still a short-period model that also
suers from some other limitations.
In recent years, perceptive public finance economists such as A.B.
Atkinson, J.E. Stiglitz, L. Kotliko, L. Summers and R.A. Musgrave have
recognized that there are serious problems with the mainstream approach
to taxation. Their concerns have centred principally on the incompatibility
of the micro and macro elements of the orthodox approach. However,
because they have constrained themselves to working within the neoclassical paradigm, they have been unable to find a solution. If the incidence and
macroeconomic eects of taxation are to be better understood, they will
have to be approached from outside the orthodox framework.
Taking a Post Keynesian approach requires us to go back to Kaleckis
path-breaking paper of 1937 and show how the introduction of taxation
modifies the various elements of his theoretical schema his theories of
income determination, income distribution, investment, business cycles and
growth. In the event, the introduction of taxation into Kaleckis theory does
not fundamentally alter the micromacro relationships that characterize his
entire approach. At the macroeconomic level, the aggregate flows in the
economy determine the level of profits. At the microeconomic level, the
degree of monopoly, as reflected in price/prime cost mark-ups, determines
the distribution of income. Tax policy can aect the aggregate flows of
spending and profits, but firms pricing decisions, as reflected in their markups of price over prime cost, determine the intra-industry and inter-industry
328
Taxation
329
and class distributions of income. Ultimately, the confluence of these factors

determines the short-period incidence of taxes. It is this incidence, through
its impact on firms investment decisions, that generates a long-period eect.
This critical interdependence between the microeconomic and macroeconomic forces in Kaleckian theory in general, and tax theory in particular,
provides a framework that is lacking in mainstream public finance theory.
In his original 1937 paper, Kalecki analysed the short-period eects of
the taxation of commodities, income and capital on employment, the level
of national income and its distribution. There is an important interdependence in Kalecki between the level and the distribution of national income.
As he stated (Kalecki 1968, p. 61): Gross income . . . is pushed up to a point
at which profits out of it are determined by the distribution factors.
Kalecki used the term degree of monopoly to describe these factors that
determine the income shares of profits and wages such as industrial concentration, product dierentiation, entry barriers or trade union power.
Using Kaleckis famous profit equation (where aggregate profits are defined
as the sum of investment, the government budget deficit, net exports and
the dierence between capitalist consumption and worker savings), the
short-period macroeconomic eects of taxes on profits or wages depend on
three things: (i) the relative marginal propensities to spend out of wages and
profits; (ii) whether or not there are compensating changes in the level of
government expenditure; and (iii) the extent to which a tax change is shifted
through a change in business mark-ups.
Post Keynesian/Kaleckian tax analysis becomes more complex when we
move out of the short period. The first step is to integrate Kaleckis theory
of taxation with his theory of the business cycle (Laramie and Mair 2000).
This shows that the taxation of wages and profits can have short-period
eects on profits and that the short-period eects of the taxation of profits
can have long-period eects on investment. These long-period eects of the
taxation of wages and profits can aect the amplitude of the business cycle
and the trend rate of investment. The impact of the tax system on the business cycle and the trend is derived from Kaleckis investment function. The
two channels through which the tax system operates are the rate of depreciation and the level of profits. The depreciation channel operates by aecting the real tax bill associated with old equipment. As this equipment
becomes physically and financially less productive as a result of technological change, an increase in the rate of tax on profits will increase the rate of
depreciation, accelerate its obsolescence and encourage its replacement by
newer, more ecient and more profitable plant. The eect on the level of
profits of a change in taxation operates directly through a change in tax
rates and indirectly through a change in the wage share of national income
on the flows of spending that comprise aggregate profits.
330
Taxation
These two eects can then be analysed at various stages in the recovery
and downswing of the business cycle. The eects of tax policy on the amplitude of the business cycle depend critically on whether or not the wage tax
and the profit tax are shifted. In the simplest scenario, we assume: (i) no
shifting of either a wage tax or a profit tax; and (ii) that the eect of a tax
change on the level of profits is greater than its eect on the rate of depreciation. Then an increase in a tax on profits will reduce profits and dampen
the amplitude of the cycle, and an increase in a tax on wages will increase
profits and attenuate the cycle. This example is only one of a large number
of possible permutations of tax change and tax shifting but it illustrates the
potential oered by a Post Keynesian/Kaleckian model for a stabilization
role for fiscal policy.
The eects of taxation on the long-period growth of the economy are
even more dicult to establish. In the first place, there must be a theory of
growth to provide a framework of analysis. The problem is that the form in
which Kalecki left his growth theory when he died was unsatisfactory, with
undue emphasis on cautious capitalism. A more balanced version of
Kaleckis growth theory has been developed by Gomulka et al. (1990).
Introducing taxation into this revised version allows us to examine the
eects of balanced changes in the structure of taxation on long-term
growth, stability and employment (Laramie and Mair 2000). As with the
eects of taxation on the business cycle, the results are complex and heavily
dependent on whether or not the taxes are shifted via changes in business
price/cost mark-ups. Nevertheless, it can be shown that there are circumstances under which balanced changes in the structure of taxation as
between wages and profits will achieve greater stability, promote faster
long-term growth and reduce unemployment.
It is not only at the aggregate level of the economy as a whole that Post
Keynesian/Kaleckian tax theory can be applied. Typically, in discussions of
fiscal policy, it is the role of central or federal government that is emphasized, while that of states or local governments is ignored. But in many
countries, state and local government receipts and expenditures represent a
significant share of GDP and thus aect aggregate spending and economic
growth. Also, much of the provision of government goods and services is
done at the state or local level and is likely to have aggregate demand and
aggregate supply eects. A Post Keynesian/Kaleckian approach to state
and local taxes seeks to identify their eects on post-tax profits. This is done
by separating out state, local and federal budget deficits and examining the
institutional and political factors that have determined the budget stances
of state and local governments. Laramie and Mair (2000) have shown that
in the United States, for example, because of institutional and political
factors, the macroeconomic eects and incidence of state and local taxes
Taxation
331
are remarkably dierent from the incidence of similar federal taxes on corporate profits.
Post Keynesian/Kaleckian analysis also provides a framework within
which to consider the macroeconomic eects of fiscal policy in the European
Union. The establishment of the European Central Bank and the introduction of the euro have resulted in a loss of national monetary autonomy that
requires adherence to new fiscal constraints. Post Keynesian/Kaleckian tax
theory suggests that stabilization need no longer depend on the application
of the orthodox Keynesian policy of running budget surpluses or deficits.
Stabilization objectives can be achieved within given volumes of government revenue and spending by altering the structure of taxation. Thus compliance by a member state of the eurozone with the budgetary constraints of
membership need not act as a constraint on its ability to use fiscal policy as
an instrument of stabilization. This is an implication of Post Keynesian/
Kaleckian tax analysis which would enable member governments of the
European Monetary Union (EMU) to retain a significant degree of autonomy over fiscal policy without violating the rules of the Maastricht Treaty.
A significant feature of Kaleckian tax analysis is that it provides a challenge to the supply-side argument that economic growth can be stimulated
only by cutting government spending and taxation. The theoretical and
empirical arguments that underpin supply-side economics are rather weak.
In particular, this line of argument is based on the marginal productivity
theory of income distribution, where the economy is constrained by
supply-side determinants. In contrast to the orthodox approach in which
the real wage rate is assumed to be determined in the labour market, a
Kaleckian approach argues that the real wage is determined in the product
market. This underlines the fundamental dierence between the Kaleckian
and the orthodox (including early Keynesian) approaches to taxation. Post
Keynesians have to recognize the political and economic constraints on the
ability of demand management policy by itself to bring an economy to the
level of full employment and keep it there. Chief among the political constraints is the one that Kalecki recognized as early as 1943, namely the
political power of rentiers and capitalists (Kalecki, 1943). The macroeconomic incidence and eects of taxation in a Post Keynesian/Kaleckian
model depend critically on whether or not tax shifting occurs. Tax shifting
occurs when either workers or capitalists are able to engineer a favourable
shift in business mark-ups in response to an increase in the taxation of
wages or profits, in order to maintain or increase their share of national
income. Business mark-ups and Kaleckis underlying degree of monopoly
theory of income distribution play a pivotal role in determining the macroeconomic outcomes of fiscal policy.
This issue poses a serious dilemma for governments in the twenty-first
332
Tax-based incomes policy
century. If governments seek to pursue fiscal objectives of stimulating

investment and growth within a balanced budget framework by reducing
the taxation of profits, they run the risk of achieving the exact opposite,
depending on how the factors that determine the degree of monopoly, and
hence the distribution of income, pan out. Governments may find themselves increasingly having to try to resolve contradictory demands from
dierent groups in society over what may be considered a socially equitable distribution of income. Post Keynesian/Kaleckian tax theory identifies
profound implications for the future of capitalism, implying that fiscal
policy can either counter or reinforce the very nature of capitalism. The
ultimate eect of fiscal policy on stability and growth depends on how
aggregate spending flows and business mark-ups react to tax changes, in
other words on how the level of profits and distribution of income are
modified.
A J. L
D M
See also:
Budget Deficits; Fiscal Policy; Kaleckian Economics; Tax-based Incomes Policy.
References
Asimakopulos, A. and J. Burbidge (1974), The short-period incidence of taxation, Economic
Journal, 84 (334), 26788.
Gomulka, S., A. Ostaszewski and R.O. Davies (1990), The innovation rate and Kaleckis
theory of trend, unemployment and the business cycle, Economica, 57 (228), 52540.
Kalecki, M. (1937), A theory of commodity, income and capital taxation, Economic Journal,
47 (187), 44450.
Kalecki, M. (1968), Theory of Economic Dynamics, New York: Monthly Review Press.
Laramie, A.J. and D. Mair (2000), A Dynamic Theory of Taxation: Integrating Kalecki into
Modern Public Finance, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Tax-based Incomes Policy

The tax-based incomes policy (TIP) was an innovative proposal that
received serious consideration in the United States from leading economists and policy makers in the 1970s in response to the severe bouts of stagflation simultaneous inflation and recession. Incomes policies, as they
were called then, tried to bring governmental pressure on leading firms to
hold down their price and wage increases, thereby reducing inflation
without utilizing the painful pressure of a deep recession. TIP was intermediate between voluntary wage-price guidelines and wage-price controls.
TIP used financial incentives to try to induce firms to hold down price and
Tax-based incomes policy 333

wage increases. Advocates of TIP felt voluntary guidelines were too weak,
controls too strong, but that TIP was just right and compatible with a
market economy that ran on financial incentives.
TIP was advocated as an alternative to unleashing central banks to
subdue inflation by a draconian tight monetary policy that worked by generating a deep recession and high unemployment, forcing desperate
workers to accept small wage increases, which led to small cost increases
and small price increases. In the early 1980s, central banks in many countries, including the United States, took this painful course. Inflation was
indeed subdued at a high social cost, and the TIP proposal receded from
debate and discussion. The past two decades have been fortunate in generally avoiding stagflation shocks in most countries, inflation has remained
generally low, and TIP has remained dormant. It is possible, however, that
some day another upward supply shock, such as the one administered by
OPEC (Organization of Petroleum Exporting Countries) in the 1970s, will
once again inflict economies with another serious dose of stagflation. If
that day comes, there may once again be renewed interest in the tax-based
incomes policy.
Although TIP came to win the support of several mainstream Keynesian
economists in the 1970s, such as Arthur Okun of the Brookings Institution
and James Tobin of Yale University, and even the co-authorship of a mainstream conservative economist, Henry Wallich, a member of the Board of
Governors of the Federal Reserve, the original proponent and most spirited advocate for TIP was the distinguished Post Keynesian economist,
Sidney Weintraub of the University of Pennsylvania (Seidman 1983).
Although Weintraubs articles proposing TIP first appeared in 1971, the
Post Keynesian theory that led him to the proposal began in the late 1950s.
In 1959 he published A General Theory of the Price Level, presenting theory
and evidence linking the money (nominal) wage to the price level, and challenging the quantity (of money) theory of inflation with an alternative
wage-cost mark-up theory of inflation. He proposed watch-tower control,
a voluntary wage guideline equal to average productivity growth. As inflation heated up in the second half of the 1960s, Weintraub continued his
work on the key role of the money wage in the inflation process, and the
need for a wage policy (incomes policy). He gradually became convinced
that voluntary guideposts were too weak. He was passionately opposed to
the use of unemployment to restrain wage inflation, but he was also averse
to wage and price controls as too rigid for a sustained solution during
peacetime. So what was left? In 1970, Weintraub proposed a method he
believed would be stronger than voluntary guidelines, but more flexible
than controls: a tax incentive. His suggestion appeared as a short piece in
the New York Times entitled, A proposal to halt the spiral of wages and
334
prices. Conservative Henry Wallich had oered a similar idea in a

Newsweek column in 1966. Weintraub then published a full-length discussion of the proposal in the British journal Lloyds Bank Review in January
1971 entitled An incomes policy to stop inflation.
Leonard Silk, economics columnist for the New York Times, brought
Weintraub and Wallich together for a luncheon. The result was the joint
article, A tax-based incomes policy, which appeared in the Journal of
Economic Issues in June 1971. By the end of the decade, after two bouts of
stagflation, TIP began to make its way into policy discussions and into the
textbooks. By the early 1980s, virtually every macroeconomics textbook
presented a brief description of the TIP proposal, citing Weintraub and
Wallich as its authors.
How did Weintraub arrive at TIP? In his 1959 book, Weintraub asserted
that price is roughly a constant mark-up, to which he gave the letter k, on
unit labour cost, which in turn equals the money wage divided by productivity. If the mark-up k stays roughly constant over time in an economy, it
follows that price will remain stable if the wage grows at the same rate as
productivity. He presented empirical evidence supporting the relative constancy of k. Weintraub went further and asserted that the wage-cost markup equation is a more useful and reliable way to analyse inflation than the
famous quantity of money hypothesis, which states that the product of
money and velocity equals the product of the price level and real output,
and that velocity V is relatively constant. Weintraub presented empirical
evidence showing that whereas the mark-up k is relatively constant, velocity V is erratic and volatile. Hence, he concluded that the money wage is a
more reliable cause of the price level than the money supply.
But what, in turn, causes the money wage to change? Most economists
believed then and now that the money wage is an endogenous variable determined by the unemployment rate and expected inflation (the augmented
Phillips curve hypothesis), or even somehow determined directly by the
money supply. Here Weintraub sets out perhaps the most controversial part
of this theory: the hypothesis that the money wage is largely exogenous.
Here is a quotation from his 1961 book, Classical Keynesianism (p. 51):
The money wage partakes of the character of a rather fully exogenous variable.
It was this view that commended itself to Keynes, who felt that the money wage
that emerged from collective bargaining did not lend itself to description in
terms of neat and determinable demand and supply curves with their implicit
recognition of the market forces.
In his 1978 book, Capitalisms Inflation and Unemployment Crisis,

Weintraub responds directly to the conventional view of what determines
the money wage, the augmented Phillips curve hypothesis (p. 107):
Tax-based incomes policy 335

According to the new disciples, W can be most anything, depending on expectations. This is a curious theory indeed to be promulgated by endogenous proponents, for the exogeneity aspect is simply demoted from money wages and
promoted in expectations.
Weintraubs focus on the money wage in the inflation process, his conviction that the wage is largely exogenous, and his opposition to the use of
unemployment to restrain wage inflation, led him in 1970 to propose TIP.
Perhaps his most complete exposition of TIP is in his 1978 book,
Capitalisms Inflation and Unemployment Crisis, where he devotes chapters
68 to its theory, design and practical implementation.
Weintraub regarded TIP as fundamentally dierent from mandatory
controls. It is important to quote from his 1978 book on this point, because
some analysts, especially critics, have characterized TIP as essentially the
same as controls. Weintraub introduces TIP as follows (pp. 1223):
Rather than emulating the methods of the collectivist societies which issue commands to do it or else, free societies must combat inflation by replicating the
incentives and deterrents of the price mechanism. An eective instrument is
already at hand in the traditional tax powers; they need not be exercised in an
onerous way, especially since revenue is not the objective. As in most good law,
the tax regulations must allow legal circumvention to impart flexibility; evasion
can be permitted but at a price, in the manner of a trac fine for conscious
speeding, maybe under emergency contingencies . . . An Incomes Policy enforced
through the tax mechanism can thus be compatible with a market economy and
the broader attributes of freedom.
Clearly, Weintraub is a Post Keynesian, not a Post-Marxian. His aim is to

fix the market system, not supplant it.
Like Keynes, Weintraub generally accepts the pattern of relative wages
and prices, and the resulting allocation of resources, generated by the market
economy. He simply wants to induce smaller wage and price increases less
inflation without interfering with the pattern of relative wages and prices
and the resulting allocation of resources. His highest priority is to avoid
fighting inflation through tight money and high unemployment.
Weintraub devotes more than two pages to the shortcomings of permanent wage-price controls, beginning the discussion as follows (p. 154):
Pervasive and meticulous price and wage controls are inimical to the functioning of the market economy. At best, they are a form of shock therapy to force a
healthier frame of mind on inflationary expectations as a temporary remission
interlude. Unless innocuously applied as a toothless bark, they must end up
diverting innumerable and mostly trivial private decisions into public forums
for tedious hearings before committees under political aegis.
In these three chapters, Weintraub devotes substantial space to practical

administration. He emphasizes that TIP should be limited to the largest
336
corporations perhaps 1,000 that set the wage-price pattern in the

economy; and that a sti tax penalty on the corporation is the key to TIPs
strength. He opposes some recent suggestions that TIP apply to most firms
in the economy and that a tax reward be used instead of a tax penalty.
One problem that has plagued the original WeintraubWallich version of
TIP is that it appears anti-labour, because it contains a wage standard but
no price standard. In their 1971 article, Wallich and Weintraub explain that
a price standard should be unnecessary according to the wage-cost markup equation, and may prove dicult to administer. In Capitalisms Inflation
and Unemployment Crisis, Weintraub oers several suggestions for a more
balanced TIP package. It is possible that a balanced price-wage TIP, limited
to large corporations, may prove able to win greater political support than
the original wage-only version, yet still prove administratively feasible.
Work on making TIP appear more balanced took place in the late 1970s
(Seidman 1978). Should stagflation return, and a revival of interest in TIP
occur, some of this work may prove useful.
Although Weintraubs work appears to focus on inflation, it is his
passion for preventing high unemployment that reveals the Post Keynesian
spirit of his work. What alarms him is the apparent willingness of many
economists and policy makers to use high unemployment as the main
weapon against inflation. In his 1961 book, Classical Keynesianism, he
wrote (pp. 3940):
If unemployment is the answer to the inflation problem, then Keynesianism as a
social philosophy is dead, literally interred by Keynesians and, curiously, all in
the name of the mentor . . . The only alternative to conscious and deliberate creation of unemployment in which the forces of aggregate demand are running
strong, and one which could still preserve the analytical and philosophical system
fashioned by Lord Keynes, would be some deliberate design and control of the
money wage level, just as we have learned to use fiscal and monetary policy to
control aggregate demand . . . How to contain the money wage level while maintaining full employment, and without breeding unacceptable interferences in
labor markets, comes close to being the most important policy issue of our day.
One decade later Weintraub (1971) published his Lloyds Bank Review article
proposing in detail a tax-based incomes policy.
L S. S
See also:
Inflation; Stagflation; Wages and Labour Markets.
References
Seidman, L. (1978), Tax-based incomes policies, Brookings Papers on Economic Activity, 2,
30148.
Third way
337
Seidman, L. (1983), Sidney Weintraub, the man and his ideas, Challenge, 26 (5), 228.
Wallich, H. and S. Weintraub (1971), A tax-based incomes policy, Journal of Economic
Issues, 5 (2), 119.
Weintraub, S. (1959), A General Theory of the Price Level, Philadelphia: Chilton.
Weintraub, S. (1961), Classical Keynesianism, Westport, CT: Greenwood Press.
Weintraub, S. (1971), An incomes policy to stop inflation, Lloyds Bank Review, 99, 112.
Weintraub, S. (1978), Capitalisms Inflation and Unemployment Crisis, Reading, MA:
Addison-Wesley.
Third Way
In the 1980s, parties of the left found themselves in a severe political crisis,
for an obvious reason: peoples expectations of the basic institutions those
political forces had helped to establish were increasingly disappointed. The
historical achievement of social democracy (the term includes similar
parties around the world), that is, the welfare state, began to show signs of
financial crisis due to overexpansion, rising unemployment and tax revolt,
as well as ineciencies reflected in rising costs of services combined
with decreasing quality of delivery. The Soviet model based on collective
property and central planning had disappointed expectations of steady
improvement in individual and collective consumption. Both the communist and the social democratic establishment had discredited themselves by
legal and illegal enrichment, though on diering scales.
The implosion of the Soviet system intensified the need for a new strategy by the democratic left. It put an end to the momentum of systems competition. The capitalist system of market economies thus permanently lost
its constraining pole. The monopolar world order of US dominance
soon became manifest. It is based on a militarytechnological, a monetary
industrial and a mediaideology core and is decisively supported by a web
of alliances ranging from the North Atlantic Treaty Organization (NATO),
the European Union and the Organization for Economic Cooperation and
Development (OECD) to the International Monetary Fund (IMF) and the
World Trade Organization (WTO), but also includes Microsoft, CNN,
Moodys and certain foundations. Neoliberal thought has become the paramount hegemonic ideology, the only one with global reach. The answer to
any political question, be it economic, social or scientific, will tend to
comply with this dominant world-view. It is very dicult, sometimes risky
and mostly in vain to argue for other possible answers. The response, in
most cases, is clear: priority is to be given to the market, particularly to
monetary stability, deregulation and the privatization of previously public
tasks and public property. Those favouring the privatization of anything
public have adequately named this the Washington Consensus.
The modernization of social democracy and the identification of a
338
Third way
post-communist perspective was attempted under several labels such as

Third Way, New Labour, Neue Mitte or Modern Governance. In this
entry, in order to avoid futile terminological exercises, the expression Third
Way is used to cover all major projects for modernizing the moderate left.
The first such attempt was initiated in the US by the Democratic Leadership
Council, a group supporting Bill Clintons presidential candidacy. For quite
some time, at least in Europe, political and intellectual initiatives by Tony
Blair (Blair 1998) and Anthony Giddens (Giddens 2000) received considerable attention. To further the integration of varying modernization currents,
the term Multiple Third Ways was introduced. (A book with this title
(Cuperus et al. 2001) gives an account of all shades, including dissenting
opinions in French and Italian centre-left parties.) In this entry, preference
is given to the expression Thirdwayism.
In political practice, Thirdwayism may be characterized by three basic
elements: (i) acceptance of monetary stabilization, deregulation and privatization as top priorities in economic and social policy; (ii) a changed role
for the state, from caretaker to empowering agency; and (iii) acceptance of
US dominance in technology and military fields. Thus due attention is
given to the core areas of US dominance. In actual policy making, the
supply side receives top priority in programmes promoting technology,
innovation and research, as well as education at all levels, including forms
of lifelong learning. Full-employment Keynesianism has thus been
replaced by labour market policy. The emphasis is now put on flexibility,
which is promoted by deregulation and training in certain skills, as well as
a reduction of unemployment benefits. Responsibility for finding a job is
individualized. Everybody becomes a manager of his or her own human
resources, individually responsible for his or her own upkeep. It is the modernized welfare states task to empower people and restrict public support
to those objectively disabled. Abuse of public welfare is to be minimized,
over-ruling the principle of free choice of employment. Inequality in
income and wealth, even if growing dramatically, is accepted as a way of
promoting economic growth, following the assumption that such growth
will eventually trickle down to benefit all. Equality is a political concern
only with respect to provision of educational opportunities and mitigation
of social exclusion.
In order to justify this change in political practice one employs stylized
facts. Top on the list are necessities imposed on all economic subjects and
political actors by globalization. This is claimed to be engendered by the
technological revolution, which demands a knowledge-based society and a
New Economy. According to this world-view, claimed to be superior to old
theories like Keynesianism, the new economy was, until the economic
downturn of 2001, regarded as invulnerable to business cycles.
Third way
339
At the theoretical level, Thirdwayism constitutes a radical break with the

past. In the economic sphere markets are conceived as self-regulating. In
the political sphere, the leftright divide is declared to be inadequate for
describing modern society. The centre is defined as radical and the establishment of consensus is declared to be the essence of politics.
Thirdwayism may be criticized regarding both practice and theory.
Practical performance, evaluated by the voter, is ultimately decisive. As
there are positive elements in any political project, in the case of
Thirdwayism this mainly applies to the envisaged reform of the state
bureaucracy and of public institutions. A turn away from a more authoritarian to a more cooperative mode of functioning is a topical problem in
all modern societies. This reform perspective is reinforced by the need to
cope with financial crises in the various public domains. A whole range of
supply-side goods, such as education, research and development must also
be included in any up-to-date reform programme. No doubt, Third Way
parties are on the right track in a number of issues. Thirdwayism, however,
lacks a theoretical basis of its own. Most of its theoretical arguments are
borrowed from mainstream social thought. This lack of theoretical
backing seriously impedes it from regaining the intellectual initiative over
hegemonic neoliberalism and reintroducing social and economic reform.
Globalization, based on technology, is also perceived in a rather pervasive way by Thirdwayism. Due to inevitable global competition, this phenomenon is seen as permeating investment, production, trade and finance,
as well as information and research. It is regarded as a natural process,
selecting the fittest modes in an evolutionary sense. The obvious fact that
globalization proceeds along a manmade path and is defined by manmade
rules remains completely out of the picture. A closer look would reveal
those rules to be construed by US players, in alliance with other global
actors. The Washington Consensus is the dominant prescription for solving
almost all economic and social problems on this globe. For example, this
set of recipes is mandatory for countries with structural payments deficits
that need to qualify for loans by the IMF and related institutions.
Thirdwayism does not even mention the Washington Consensus or IMF
conditionality. In short, Thirdwayism boils down to an attempt to adapt
the reasoning and action of former left-wing and centre-left parties to the
ideological hegemony of neoliberalism.
Markets are by no means self-regulating, but in a systematic way unstable and open to failure. Performance which is socially acceptable, therefore,
requires regulation and active policy intervention. In its radical version,
Thirdwayism accepts that states are embedded in markets. In a similar vein,
there is no proof that private property is generally preferable to other forms
of ownership. Given this, economic policy inspired by J.M. Keynes and
340
Third way
Michal- Kalecki or Post Keynesian economics cannot be dismissed as theoretically outdated and politically impracticable. Thirdwayisms antiKeynesianism results from submission to the dominant microeconomic
approach, which considers macroeconomics in general, and an active fiscal
and monetary policy in particular, as being erroneous and futile. This
judgement, however, is based on insucient analysis. Thirdwayism rejects
Keynesian methods of increasing private consumption through redistribution of income and increasing government expenditure on social consumption goods, in particular if they are financed by budget deficits. In line with
the Washington Consensus, only boosting investment through tax reduction and expenditure in infrastructure is admissible, even if this in fact is
also one of the main ways to increase employment in Keynesianism. By
depicting that school of thought as outdated, circular flow reasoning is also
removed from the analytical arsenal. (This is comparable to abolishing
double-entry accounting as an analytical tool in business!) The political
consequences are severe. First, it is no longer possible to assess whether the
promotion of private investment and exports is based on sucient aggregate demand. Second, eective demand shortages, for example, those due
to petrol-price inflation or the asymmetric adaptation mechanism, which
limit eective demand in both debtor and creditor countries, are overlooked. Third, asset-price inflation (or deflation), so important for understanding global stock exchange movements during the years preceding and
following 2000, becomes inexplicable. Last but not least, economic instability and ensuing avoidable damage can no longer be analysed in terms of
institutional deficiencies, as they are instead primarily attributed to individual misbehaviour, such as inflexibility or welfare dependency.
If the leftright divide is correlated with more or less equality in income
and wealth, or access to opportunities and power, it is hard to see why this
distinction should be regarded as outdated. Given the dramatic increase in
inequality in recent decades, it is equally dicult to comprehend how consensus as a primary virtue can play such an overwhelming role in any
important up-to-date political project. Indeed, in advanced political theory
the dierence between us and them is being reactivated as a centre-piece
of politics. It allows one to define the dividing line between political adversaries and to mobilize supporting interests.
Thirdwayism can, finally, be characterized by its neglect of three major
issues:
1.
What is the modern public purpose? A redefinition of the public

purpose is pertinent at the local, regional, national, continental and
the global levels. It is a precondition for defining the adversary and for
mobilizing supporting interests.
Time in economic theory 341

2.
3.
Which new role should be ascribed to public property? The redefinition of property, perceived as a bundle of rights, reflecting (a role for)
the private and the public interest, is an inevitable theoretical task. Its
urgency is exemplified by at least two phenomena: first, by the
public-good quality of old and new knowledge; second, by WTO regulations allowing the privatization of intellectual property, such as
scientific discoveries.
What role should institutional diversity play and what forms might it
take? Institutional diversity and awareness of cultural legacies have
been found to constitute a comparative advantage in economic development. Their neglect was one of the reasons for the Soviet Unions
inferiority in systems competition. By elaborating on those issues,
strangely ignored by Thirdwayism, an agenda for the future could
take shape.
This is, however, not achievable without overcoming the ideological hegemony of neoliberalism. This endeavour, if successful, opens up the path to
a post-Washington Consensus, or, formulated positively, to an agenda for
a future global res publica.
E M
See also:
Economic Policy; Globalization; Transition Economies.
References
Blair, Tony (1998), The Third Way: new politics for the new century, London: Fabian
Pamphlets, 588.
Cuperus, Ren, Karl Duek and Josef Kandel (2001), Multiple Third Ways. European Social
Democracy Facing the Twin Revolution of Globalisation and the Knowledge Society,
Amsterdam, Berlin and Vienna: Wiardi Beckman Stichting et al.
Faux, Je (1999), Lost on the Third Way, Dissent, Spring, 6776.
Giddens, Anthony (2000), The Third Way and Its Critics, Oxford: Polity Press.
Laclau, Ernesto and Chantal Moue (2001), Preface to the second edition, in Ernesto Laclau
and Chantal Moue, Hegemony and Socialist Strategy. Toward a Radical Democratic
Politics, London: Verso, pp. xiixix.
Matzner, Egon (2000), Monopolar World Order. On the Socioeconomics of US Dominance,
Szombathely: Savaria University Press.
Time in Economic Theory

Conventional (neoclassical) economics allows no place for time in the sense
of a historical, uni-directional process. While Alfred Marshall observed
that time was the source of many of the greatest diculties in economics
(Principles of Economics, 8th edn, p. 92), neither he nor his descendants
342
Time in economic theory
were able to incorporate meaningful time into their theory. Equilibrium

analysis, whether of a partial or a Walrasian form, cannot manage time in
its historic context. Contrary to textbook authorities, the short and long
runs of neoclassical theory do not refer to the passage of historic (calendar) time, but are mere analytic concepts that allow time enough for whatever logically needs to happen to permit the equilibrium outcomes intrinsic
to the argument to unfold. In the general equilibrium world, time is essentially excluded from the theory because everything happens simultaneously
there is no process or sequence of events taking place over real time.
Various attempts have been made to deal with time in a more realistic
fashion, but still retain the notion of equilibrium states to which the
economy is always tending. Such eorts, since they are contained within an
equilibrium framework based on a Cartesian deductive method, must
reduce time to some non-historic meaning: economic processes must be
reversible and the future must be knowable if equilibrium outcomes are to
have any meaning. As the neoclassical equilibrium argument is of a
mechanical nature, similar to the inner workings of a clock-like machine,
one can trace the sequence of events through a system of interconnected
variables (gears) in any direction: a deterministic outcome will always
obtain and no arrow of time is permitted by the argument. Hickss weeks
are not to be confused with the passage of real time (Hicks 1939). Week 2
does not historically follow week 1, because the equilibrium adjustment
process undergone in week 1 could just as logically follow that of week 2:
weeks are interchangeable. And, if processes are reversible and the future
knowable, any expectations admitted in this theory (if the assumption of
full information is dropped) must also be reversible. There can be no uncertainty in the sense that we simply do not know what the future state(s) will
be (Dow 1996, chapter 6).
The notion of equilibrium is further constrained by the requirement that
economic laws be of a universal nature, applicable to all places and all
times. This connotes that, although superficial appearances seem to indicate that the world undergoes change, no fundamental change is possible.
Since change can take place only through the passage of time, the neoclassical theory cannot incorporate history.
For mainstream theorists, time is simply an intellectual construct developed to conform to the equilibrium requirements of the theory itself.
Assumptions of perfect knowledge of past, present and future states
where there are no disappointed expectations in the past and where the
future is known emasculate the very concepts of time and uncertainty in
economics (Davidson 1978, p. 14). In the bounded rationality version of
the argument, the precise future is not known, but all possible future states
are knowable. Economic agents can still behave rationally on the basis of

quantifiable probability estimates of the future, though suboptimal (satisficing rather than optimizing) results will prevail. This is not because the
future is yet to be created (implying the passage of historical time), but
because decision makers are constrained by their own cognitive limitations.
Agents intend to behave rationally in the sense of optimizing some variable
(utility), but cannot because they are incapable of acquiring the knowledge
to do so and do not have the necessary computational skills that would
allow them to use such knowledge in an optimizing calculation. The fundamental problem, then, lies not with an uncertain future, and thus with
time, but with the cognitive limitations of the economic agent. Were it not
for these limitations, agents could optimize because they could be certain.
But they must be content with a limited rationality as they can only assign
a probability to the success of any particular decision. This approach could
be termed probabilistic determinism rather than the certain determinism
of the more traditional theory (Dunn 2001).
For Keynes, and those economists who follow his lead, time is of crucial
importance in understanding the workings of a monetary (capitalist)
economy, and is inextricably linked to fundamental uncertainty and to
money itself. Economic processes occur in real time, where actions taken
today that cannot be undone have unknowable consequences for the future.
This position contains the essential Post Keynesian criticism of neoclassical equilibrium analysis: since adjustment processes necessitate change, and
change can take place only in real time, by the time the logical process
leading to equilibrium could work itself out, the conditions on which the
equilibrium outcome was specified have all been altered: no deterministic
outcome is possible: economic decisions are made by human beings facing
an uncertain and unpredictable future while they are moving away from a
fixed and irreversible past (Davidson 1991, p. 32). For Keynes, neoclassical
equilibrium analysis was just a pretty, polite technique which tries to deal
with the present by abstracting from the fact that we know very little about
the future (Keynes 1936 [1987], p. 115). For some purposes Keynes did
employ a concept of logical time, but such usage was confined to countering standard theorys arguments on its own terms or in establishing preliminary statements identifying various causal relations; this is of little interest
in analysing his general theory, which is conducted in historical time (see
Robinson 1980).
Keyness position on money and uncertainty requires a theoretical stance
on time, where economic processes occur in real time and within a distinct
form of economic organization that cannot be analysed on the basis of universal laws deduced from a reductionist (individual agent) foundation.
Production processes obviously occur through time, but this fact, while
important for Keynes, was not the defining aspect of a monetary economy.
344
Time in economic theory
Production occurs in all forms of economic organization, but a monetary

economy diers in its fundamental nature from non-monetary economies.
In a monetary (capitalist) economy, the objective of production is to
produce not goods, but money. In his early drafts of The General Theory,
Keynes, following Marx, was very specific in his formulation of such an
economy as M-C-M, where money is required to purchase inputs to
produce commodities that are then sold for money. The objective, of
course, is that the money at the end of the circuit (M) is greater than that
at the beginning a profit is made.
For Keynes and the Post Keynesians, such an economy is driven by aggregate spending. If current spending is limited to income received in the previous period, spending cannot increase beyond the level of that period M
could not exceed M. As capitalist production is undertaken primarily on the
expectation that M will grow, such an economy requires a source of spending in excess of current income debt (and its dialectical opposite, credit).
Indeed, the production process in a monetary economy must begin with
debt, because workers must be paid and capital goods purchased before
income-yielding output is produced. Hence the owners of productive property must not only incur debt to allow production, but must also hope that
sucient spending will occur in the future so that M exceeds M and debt
can be cleared (or at least serviced). The capitalist market, then, is not a
device to clear goods through changes in prices, but a process to clear debt.
And, as a lender makes loans today on the promise of payment in the future,
this process requires the creation of dated contracts; these in turn require the
establishment of a unit of account that allows the recording of debt:
A money of account comes into existence along with debts, which are contracts
for deferred payment, and price lists, which are oers of contracts for sale or purchase. Such debts and price lists . . . can only be expressed in terms of a money
of account.
Money itself . . . derives its character from its relationship to the money of
account, since the debts and prices must first have been expressed in terms of the
latter. . . . Money proper . . . can only exist in relation to a money of account.
(Keynes 1930 [1971], p. 3)
But, since the future cannot be known, it is impossible to make accurate

predictions as to the outcome of any production-exchange process. Hence,
in a monetary economy, production cannot be undertaken on the basis of
a known, quantifiable (or even probabilistic) calculation, as is supposed in
neoclassical models. Animal spirits (or spontaneous optimism) drive
investment and production in a world where historical time is of real importance and thus fundamental uncertainty is a fact of life.
In a monetary economy, the owners of productive property have obviously separated that property from the control of the larger community.

Since capitalists produce goods only in the expectation of future gain, the
community cannot depend on such individuals for its economic well-being.
Nor can the capitalist depend on the community should he or she encounter economic (financial) diculties: the capitalist faces an uncertain future.
To protect themselves against the vagaries of this uncertain future, capitalists must accumulate stocks of wealth. These stocks will not be in the
form of real goods (use values), for such forms of wealth are not readily
convertible into other goods (they are not very liquid), and they entail large
carrying costs. Thus, the capitalist will hold wealth in the form of money.
Since this appears objectively irrational (one cannot eat money), such
actions can only be comprehended within an economic organization where
time is important and the future is unknowable. [T]he importance of money
essentially flows from its being a link between the present and the (uncertain)
future (Keynes 1936 [1973], p. 293; emphasis in original).
For Keynes, then, historical time is one of the core features of his
general theory. It is only through the inclusion of time that one can understand the uncertainty that is central to the decision-making process under
capitalism, and it is only through an analysis of uncertainty that one can
understand the nature of money. And money, of course, is the central
characteristic of a monetary economy. In Hyman Minskys succinct words,
[o]nce a financial perspective is adopted, time cannot be interpreted away
as just adding additional commodities (Walrasian money) to the economy
(Minsky 1982, p. 62).
J F. H
Note Portions of this entry are based on J. Henry and L.R. Wray, Economic
Time, Working Paper No. 255, New York: The Jerome Levy Economics
Institute, 1998.
See also:
Agency; Equilibrium and Non-equilibrium; Keyness General Theory; Non-ergodicity;
Uncertainty; Walrasian Economics.
References
Davidson, Paul (1978), Money and the Real World, 2nd edition, New York: Wiley.
Davidson, Paul (1991), Controversies in Post Keynesian Economics, Aldershot: Edward Elgar.
Dow, Sheila (1996), The Methodology of Macroeconomic Thought, Cheltenham, UK and
Dunn, Stephen (2001), Bounded rationality is not fundamental uncertainty: a Post Keynesian
Hicks, John R. (1939), Value and Capital, Oxford: Clarendon Press.
Keynes, John Maynard (1930 [1971]), A Treatise on Money, in The Collected Writings of John
Maynard Keynes, Vol. 5, edited by D. Moggridge, London: Macmillan for the Royal
Economic Society.
Keynes, John Maynard (1936 [1987]), The General Theory of Employment, Interest and Money.
346
Tobin tax
Reprinted as The Collected Writings of John Maynard Keynes, Vol. 14, London: Macmillan
for the Royal Economic Society.
Minsky, Hyman (1982), Can It Happen Again?, Armonk, NY: M.E. Sharpe.
Robinson, Joan (1980), History vs. equilibrium, in J. Robinson, Collected Economic Papers,
Vol. 5, Cambridge, MA: MIT Press, pp. 4558.
Tobin Tax
In his 1972 Janeway lecture at Princeton, James Tobin (1974) proposed a
tax on foreign exchange transactions as a way of limiting speculation,
enhancing the ecacy of macroeconomic policy, and raising tax revenues
(Davidson 1997 opposes this view). Keynes, in the Treatise on Money and
in the General Theory, had already suggested that a tax on foreign lending
to contain speculative capital movements might be necessary (see also Haq
et al. 1996). Ocial interest in the Tobin tax has been repeatedly expressed.
The United Nations Development Programme (1994) emphasized its
potential for raising large amounts of money that could be used to finance
development. The Tobin tax has been gaining popularity since then, and
governments have either shown approval or willingness to discuss it.
In Arestis and Sawyer (1997) four sets of rationale are advanced in
support of the Tobin Tax. In a world of floating exchange rates, the large
volume of transactions is viewed as generating volatility in the exchange
rate, with consequent detrimental eects on the real economies. The second
rationale is simply its revenue potential. Tobin (1978) suggested it as a
byproduct of a financial transaction tax, not as the main aim a financial
transaction tax of 0.05 per cent could have raised $150 billion a year over
the 19952000 period. The third rationale concerns the possibility of
enhancing the autonomy of national economic policy, and reducing the
constraints on such policy imposed by financial markets. In this context a
financial transaction tax increases the independence of policy makers by
reducing foreign exchange rate volatility through hitting the most frequent
transactors the hardest. The fourth is that this tax can potentially tackle the
problems just alluded to more flexibly than the introduction of financial
controls, especially quantitative exchange controls, which are usually viewed
as rigid.
Three reasons for the increased interest in the Tobin tax are mentioned in
Arestis and Sawyer (1997). The first is the growing volume of foreign
exchange trading. The second reason is that a transactions tax is now seen
as important not merely by policy makers and others concerned with foreign
exchange market volatility, but also by those who attach significance to
public financing of world development. The third reason is an increasing
realization that foreign exchange markets do not operate in the ecient
Tobin tax
347
manner portrayed in the rational expectations literature. Foreign exchange

markets suer from asymmetric information, herd behaviour, moral hazard
and from the possibility of multiple equilibria; the implication of all these
is persistent misalignments and unstable exchange rate regimes.
One common argument raised against the Tobin tax relates to its possible
distortionary eects. The argument is that such a tax leads in a competitive
market to an equilibrium that involves a lower quantity of transactions,
and fewer resources being allocated to that particular market. There are
three points to be made in connection with this argument. First, the financial sector may at present be relatively lightly taxed, and the products of
the financial sector are generally not subject to either general sales or valueadded taxes or to specific excise taxes and the like. This would mean that
the imposition of a financial transaction tax might in eect be removing
some distortions rather than imposing them. To the extent that this view is
accepted, the introduction of a financial transaction tax would help to
reduce the distortionary eect of the tax system.
Second, the distortionary nature of a tax arises from some potentially
beneficial trades not taking place that would otherwise have happened.
This leads to the question of whether there are gains from the current
volume of exchange transactions, which would not arise with a substantially smaller volume. Some further doubt is cast on the distortionary argument by the observation that while the two parties to a foreign exchange
trade may believe that they will gain from the trade (through a favourable
price movement), both cannot do so, that is it is a zero-sum game.
Third, the analysis of distortions is an equilibrium one, and it is equilibrium trades which are discouraged. But there is a sense in which much of
the trading in currency markets is disequilibrium trading in terms of
seeking to take advantage of price changes. Thus the conventional analysis
of distortions does not apply to this situation, and if it is the case that the
amount of noise trading is excessive, then a tax is beneficial rather than
distortionary.
Assuming that the tax was fully passed on to consumers, it is clear that
there could be a substantial impact on the volume of international trade
(Davidson 1997). However, a rather low rate of taxation and the tax being
absorbed by producers rather than passed on to consumers are distinct possibilities. In evaluating the overall balance of eects of a financial transaction tax on international trade, due consideration would need to be given
to the eects of reduced volume of exchange transactions, of reduced volatility, enhanced independence of national economic policies and the probable stimulus to world-wide aggregate demand. These latter factors would
stimulate international trade, and the overall net eect of such a tax on
international trade cannot readily be predicted.
348
Tobin tax
It is expected that the introduction of a Tobin tax would be a major economic and political development. At the same time it would have to be
introduced on a big bang basis, for otherwise foreign exchange dealings
would quickly move to those countries that were not applying the tax.
It is also expected that a financial transaction tax would have a significant
impact on world-wide aggregate demand. Besides the obvious point that the
aggregate demand eects will depend on the use to which the tax revenue is
put, it is quite reasonable to assume that a financial transaction tax would be
levied on those with a low propensity to spend, and the redistribution would
be towards those with a much higher propensity to spend. Hence aggregate
demand may well increase. This would be an additional eect to the
enhanced capability of national governments to pursue economic policies
that stimulate a higher level of demand. Furthermore, to the extent to which
the tax revenue does not lead (at the national and/or international level) to
increased government expenditure, there would obviously be some reduction
in budget deficits. Some would argue that the reduction in budget deficits
would lead to a reduction in interest rates, with some stimulus to investment.
The appropriate definition of the transaction should be as follows: any
transaction that involves the exchange of a financial asset denominated in
one currency for a financial asset denominated in another currency. This
was Tobins initial suggestion when he wrote:
[T]he tax would apply to all purchases of financial instruments denominated in
another currency from currency and coin to equity securities. It would have to
apply . . . to all payments in one currency for goods, services, and real assets sold
by a resident of another currency area. I dont intend to add even a small barrier
to trade. But I see o-hand no other way to prevent financial transactions disguised as trade. (Tobin 1978, p. 159)
A number of proposals have been put forward on the way to distribute

the tax proceeds (see United Nations Development Programme 1994). To
the extent that it is the International Monetary Fund (IMF) or World Bank
which are the intermediate recipients, a further proposal may be to enhance
the lending capabilities of these institutions especially to the Third World
countries, which could embrace development and anti-pollution projects.
The workings of the tax could be reinforced by making the administration
of a Tobin tax a condition of membership of the IMF. It should be conceded, though, that this may not be sucient to prevent the growth of oshore dealings. This is so since a small country would have very little to gain
from membership of the IMF as compared with the potential revenue as
the location of oshore financial markets.
There is widespread agreement that the tax would have to be implemented on a coordinated international basis. It may not be necessary for
Tobin tax
349
there to be full agreement over the tax rate, though there would be strong
pressures towards a degree of uniformity (and probably a requirement for
a minimum rate to avoid competitive undercutting of the tax rate between
countries). It is clear that there would be very considerable dierences in
the amount of tax collected in each country. Part of the international agreement could clearly be that a proportion of the tax collected be paid over to
an international body and/or used for agreed development and environmental purposes. The obvious diculty which arises here is obtaining international agreement over the introduction and the rate of the tax when the
revenue from the tax would be so unequally distributed across countries
(and to the extent to which countries fear that their financial centres would
be reduced in size, the costs also unequally distributed). Furthermore, a
substantial retention of revenue at the national level obviously reduces the
funds available for international development and global environmental
purposes. It should also be expected that the economic and political influence of the financial markets would be much reduced; indeed the imposition of such a tax would be a clear signal that the influence of the financial
sector was in decline. The point should be made, though, that the high-yield
UK securities transaction tax, known as stamp duty, has been maintained
in one of the most sophisticated financial markets. It has not obviously met
with any serious opposition from these sophisticated financial markets, nor
have there been cries of tax evasion in a market where players are most
likely to find mechanisms to evade such a tax.
Most advocates of a financial transaction tax recognize that it would
have to be universal and uniform. This requirement may well be the most
important practical obstacle to the implementation of such a tax. It would
clearly require the cooperation of all countries with significant foreign
exchange dealings within their borders, although there would be incentives
for countries to apply a lower tax rate within their jurisdiction. Given the
IMFs considerable expertise in international financial markets, it would be
in a good position to undertake such a task. Furthermore, recognition of
the IMFs central objectives of the promotion of international monetary
cooperation, to maintain exchange rate stability and orderly exchange
arrangements among its members, substantially strengthens the argument
that the IMF should play a central role in its implementation.
A Tobin tax is expected to diminish the volatility of exchange markets and
raise in revenue substantial sums of taxation. Its introduction could face
political problems; and yet it has been gaining popularity, including with the
Commission of the European Union. It might be more appropriate to use
the Tobin tax as one of several policy instruments to combat speculation on
the worlds foreign exchanges, and to finance development.
P A
350
Transition economies
See also:
Development Finance; Economic Policy; Exchange Rates; Globalization; International
Economics.
References
Arestis, P. and M. Sawyer (1997), How many cheers for the Tobin financial transaction tax?,
Davidson, P. (1997), Are grains of sand in the wheels of international finance sucient to do
the job when boulders are often required?, Economic Journal, 107 (442), 67186.
Haq, M., I. Kaul and I. Grunberg (eds) (1996), The Tobin Tax: Coping with Financial
Volatility, Oxford: Oxford University Press.
Tobin, J. (1974), The new economics one decade older, The Eliot Janeway Lectures on
Historical Economics in Honour of Joseph Schumpeter, 1972, Princeton: Princeton
University Press.
Tobin, J. (1978), A proposal for international monetary reform, Eastern Economic Journal, 4
(34), 1539. Reprinted in J. Tobin, Essays in Economics: Theory and Policy, Cambridge,
MA: MIT Press, 1982, pp. 48894.
United Nations Development Programme (1994), Human Development Report 1994, New
York and Oxford: Oxford University Press.
Transition Economies
The term transition economy can, in general, be applied to any economy
moving from one dominating mechanism for coordinating its economy to
another. However, since 1989 the term has been applied specifically to the
post-communist, centrally planned economies of Central and Eastern
Europe. Before 1989, these economies were, to a greater or lesser extent,
characterized by a high degree of administrative planning, absence of the
market mechanism in the allocation of resources, and a relative scarcity of
consumer goods as compared to the advanced Western industrialized economies.
The appearance of various ineciencies and distortions put pressure for
some sort of change in the way these economies functioned. This pressure
began intensifying after the 1970s, when the areas previously high growth
rates could no longer be maintained. Since then, and up to the time of
writing, although most of these economies have professed a commitment
to adopting the market mechanism and the incentive of the profit motive
as the only way to improve their economic performance, the record of
achievement is mixed. Ten years after the breakup of the Soviet System,
what is perhaps most striking about the European and Central Asia (ECA)
region is its diversity. A decade ago, all countries in the region seemed to
face similar challenges of transition from a planned centralized system to
a market economy (Europe and Central Asia: Introduction, World Bank
2000, p. 63).
351
The explanation for this diversity can best be explained in terms of the
presence or absence of those supportive, usually non-economic, institutions that make market mechanisms work, and market-oriented behaviour
feasible. Hence, those Eastern European economies such as Hungary,
Poland, the Czech Republic and Slovenia, which had already more diverse
economic structures, have fared better in the post-1989 years, while others,
most notably the Russian Federation and Ukraine, have experienced negative growth over much of this period, and are struggling to regain levels of
gross domestic product reached before the Soviet collapse (ignoring for the
moment the question of the reliability of pre-1989 economic statistics).
In response to an emerging recognition of various economic problems
connected with their poor economic performance, some of the socialist
economies had already begun to reduce the role of bureaucratic decision
making, starting in the 1980s. Originally, in the inter-war period, bureaucratic planning had been instituted as a way to jump-start the industrialization process; after the Second World War, it was extended to the countries
falling within the Soviet sphere of influence. Some of the characteristics of
these economies were as follows.
Planners gave preference to the production of investment and intermediate goods, and energy production, because they aimed to catch up to the
Western industrialized economies as fast as possible. Enterprises were operating entities but not decision-making units, being responsible only for fulfilling the plans targets with the resources allocated to them. Planning was
taut, targets were set high, all resources were fully employed, and there was
little slack. A price system existed, but as an accounting device rather than
as an incentive or allocative mechanism; firms did not need to respond to
financial incentives. At first, prices were more or less in accordance with
resource (opportunity) costs, but diverged considerably over time: planners
were slow to respond to changing supply conditions or wanted to manipulate the numbers for various political purposes. The result was prices that
were out of alignment with both opportunity costs and world prices for
similar items.
Various distortions and imbalances appeared, but the structure and functioning of these economies prevented their elimination. Because enterprises
were subject to a soft budget constraint meaning that they could manipulate
external financing and/or negotiate taxes, grants and credit in order to avoid
bankruptcy indefinitely there were no penalties for failure (Kornai 1980).
Hence even though shortages and bottlenecks were common, and investment
plans to eliminate them were therefore justified, they could not be fulfilled
because of resource limitations. One example was the ubiquitous existence of
queues, evidence of shortages of consumer goods. In other words, these economies were high investment, resource-constrained economies, rather than
352
(eective) demand-constrained economies, as Western capitalist market economies are. However, innovative responses were not forthcoming, because
enterprises, which had to deal with the problems, were not decision-making
units, and enterprise managers responded to political, not economic, stimuli.
In any event, enterprises had no incentive to respond innovatively, for fear that
plan targets would be revised upwards.
Whether or not the changes that were already beginning to take place
would have eventually resulted in more eective economic performance
became an irrelevant issue following the events of 198990. At that time,
the collapse of communist-dominated governments across the region initiated both political and economic changes towards the adoption of democratic political processes and the market mechanism.
Economic advisers to many of the new governments favoured a threeprong approach to economic restructuring: eliminate planning in favour of
a market mechanism; privatize productive assets to replace government
ownership; and rely on a price system to coordinate economic activity so
that financial incentives replace quantitative targets. While recognizing that
there could be some short-term costs of adjustment for example, a rise in
unemployment as enterprises got rid of excess workers, or balance of payments deficits as Western imports poured in following the opening up of
these economies advisers recommended speed, the so-called shock
therapy approach.
The assumption that markets are the best mechanism for coordinating
economic activity underlay these recommendations. In turn, this assumption underlies the standard model that focuses on self-interested individuals
making rational choices so as to maximize utility/income/welfare/profits. In
the transition economies, enterprises will have a crucial role to play in both
generating and responding to price signals because they will become financially self-sucient and therefore subject to a hard budget constraint with
real penalties for failure, rather than simply carrying out the central bureaucracys orders.
Unfortunately, and partly in explanation of the problems facing the transition economies since then, marketization is not a universal solution.
Because the standard model was devised with the already-mature market
capitalist economies in mind, it assumed that all the necessary institutions
and behaviour patterns that make the price system work were in place. Such
assumptions cannot be made in the transition economies; instead, an
approach that takes into account their history, culture, institutional development and values is called for. This alternative approach also recognizes
that it takes a long time for these supportive institutions to take root.
For example, private ownership of productive resources will only
produce socially favourable results if a legal system specifying property
353
rights, and the rights and responsibilities associated with them, is in place.
In these countries, the necessary legal changes are still being put into place
(and in some cases, required a first step of legitimizing the very concept of
private property). Enterprises can only respond to financial incentives if
there is a well-developed financial system: banks and other financial intermediaries to replace the governments role as finance provider. This further
requires the establishment of more mechanisms to enforce accountability,
including proper accounting procedures, bankruptcy laws and so on. Using
prices presumes not only that there is some way of estimating the relation
between opportunity costs and market prices, but also that there is a monetary institution capable of ensuring adequate liquidity in the system.
Replacing planning with the market mechanism assumes that behaviour
patterns are appropriate, and economic agents will respond in the right
way: for example, that enterprise owners and managers have the specific
knowledge and training to respond to price signals, and that the incentive
system does in fact harmonize the actions of consumers, employees and
managers in an acceptable way. But desired behaviour patterns are not
inherent but learned, so that the mechanisms necessary for teaching and
learning them have to be put into place.
Of particular importance to the transition economies is reworking the
role of the state. Market democracies rely on a strong eective state in order
to work as well as they do: to provide necessary infrastructure, guarantee
property rights, provide a monetary system . . . the list could go on. Also,
because of the absence of an automatic self-correcting mechanism in a capitalist market economy, there is an important role for government to ameliorate market failures and maintain the level of eective demand as a
growth-enhancing policy. This is vital once the dynamic role of (privately
generated) investment in a world of uncertainty, and in an uncertain global
economy, is recognized.
In general, the transition economies have more to do than simply replace
bureaucratic decision making and material balances planning with the
profit motive, private property and the market mechanism. They also have
to put into place all the institutions and processes that make a private enterprise economy function, as well as those that are needed for when it fails to
function properly.
Policy options compatible with a Post Keynesian approach start from the
(often dicult) realization that improved economic functioning depends on
eective operations in many dierent areas. It is not simply a question of
removing bureaucratic intervention and leaving markets alone so that the
price system can work its magic.
Certain key elements of policy can be distinguished. First, the necessary legal, social, political and economic institutional arrangements that
354
coordinate and promote correct economic behaviour should be put into

place. Those with a special impact on economic activity include financegranting institutions, with some sort of control in order to determine that
an adequate amount of liquidity exists. Money and monetary institutions
are definitely not neutral, and do influence economic activity.
Second, those elements contributing to eective behaviour patterns
should be developed. These include the dierent levels of formal and informal educational institutions, information gathering, analysis and dispersion, and control and accountability mechanisms (including accounting
rules and procedures).
Third, specific economic policy options need to be developed because a
market system does not automatically correct its failures. Hence mechanisms to maintain aggregate demand and employment at appropriate
levels, and to stabilize prices, are needed. Also, the dynamic aspects of economic growth and development are too important to be left to chance or
opportunism. This requires giving attention to investment its level and
distribution in order to promote growth in the face of an uncertain future.
Also appropriate here would be a social safety net, not only to provide
income security for its traditional clientele the retired, sick, or destitute,
for example but also to encourage risk-taking and innovative behaviour
by cushioning the hardships of failure.
Clearly, many other specific aspects of policy could be identified and
described. What is important is the flexibility that comes with the recognition that each society has its own past, present and future, and that the goal
of a pluralist, democratic market system will not be reached through a rigid
commitment to a single ocial blueprint.
C R
See also:
Economic Policy; Institutionalism; Socialism; Third Way.
References
Kornai, J. (1980), The Economics of Shortage, Amsterdam: North-Holland.
Rider, Christine (1993), The pricing problems of Eastern Europe, in Ingrid Rima (ed.), The
Political Economy of Global Restructuring, Vol. I, Aldershot: Edward Elgar, pp. 8699.
Rider, Christine (1996), Ethical policy making in the transition economies, in Edward J.
OBoyle (ed.), Social Economics: Premises, Findings and Policies, London and New York:
Routledge, pp. 17890.
Rosser, J. Barkley Jr. and Marina V. Rosser (1996), Comparative Economics in a Transforming
World Economy, Chicago: Irwin.
World Bank (2000), Annual Report 2000, International Bank for Reconstruction and
Development, Oxford: Oxford University Press.
Traverse 355
Traverse
The traverse defines the movement of the economy outside equilibrium. It
plays a particularly important role in Post Keynesian theory, as most Post
Keynesian economists have serious doubts about the relevance and usefulness of equilibrium analysis.
Economics and political economy have almost always relied on some
concept of equilibrium as a central organizing concept. A major methodological dierence between dierent schools of thought has been the operational significance that each ascribed to that concept. For some economists,
the concept of equilibrium is important for organizing ideas, and as an
idealized point of reference. Others see it as being descriptive, with actual
economies showing strong tendencies towards equilibrium, which might be
achievable except for constant shocks. However, despite the importance of
the concept of equilibrium, little was usually said about the process whereby
the economy achieved it. For equilibrium to serve the function which economists have for so long assigned to it, there must be forces pushing the
economy towards equilibrium, and the path the economy takes towards that
equilibrium, that is, its adjustment path, must not influence the equilibrium
to which it is tending. In the absence of these conditions, analysis of equilibria, independent of the traverse become pointless. Hicks, who first introduced the term traverse into economics, characterized it as the path which
will be followed when the steady state is subjected to some kind of disturbance (Hicks 1973, p. 81). In other words, the traverse describes the
economys dynamic out-of-equilibrium adjustment path in historical time.
The importance of historical time is its uni-direction, time can only move
forward, with the link between time periods given by the stock of capital
inherited from the past, and the expectations embodied in it (Robinson
1974; Setterfield 1995). Although initially the traverse was used to describe
the path between equilibria, later the traverse itself was seen to be the end
of the story. Post Keynesian economists, among other heterodox economists (especially evolutionary and institutional economists), have been
vocal about the fruitlessness of studying the equilibrium properties of an
economic system without considering the question of whether the economy
will actually get there. In other words, they have voiced their doubts about
the comparative static method which dominates modern economics. More
pointedly, Joan Robinson often criticized the separation of equilibrium
analysis from the analysis of the traverse, as she believed that the actual
equilibrium which an economy achieves (if it is capable of achieving one)
will be vitally dependent on the path it takes, so that equilibrium would
always be path-dependent (Robinson 1974).
The traverse is of relevance both to economists who deny that the
356
Traverse
economy is attracted to any equilibrium, as well as to those who accept that

the economy will tend towards equilibrium, but argue that the final equilibrium position is path-determined, in the sense associated with hysteresis
where current outcomes are determined by past values. Hicks and Lowe
undertook detailed analysis of the adjustment paths economies take
outside equilibrium. They considered the question of whether the market
would send the correct signals to allow the structure of production to
adjust as a response to a shock. The necessary adjustment path requires
both time and costs, and faces diculties which arise from disproportions
between sectors and misleading market signals (Hagemann 1992, p. 235
italics in original).
Hickss initial analysis of the traverse utilizes a two-sector, fixedcoecients model making use of the methods of the classical economists,
with similarities to Piero Sraas model. In a two-sector model the one capital
good can freely be moved between the capital- and consumption-goods
sectors. Without the complications implied by structural disproportionalities, Hicks concludes that a full-employment path to equilibrium is only
possible if the consumption-goods sector is more mechanized than the
capital-goods sector. Even if this condition is fulfilled, a full-employment traverse is not guaranteed but must satisfy a series of technologically determined
conditions with respect to the man/machine ratios in the two sectors. Hicks
reaches the important conclusion that smooth adjustment may not be possible, with prices providing inappropriate guides to decision makers (Hicks
1965). In his later work, Hicks moved away from the classical traverse, and
attempted to analyse the traverse within a hybrid neoclassical/Austrian
framework. In order to get a unique and unambiguous period of production,
he resorted to the uninteresting case of the simple profile, for reasons related
to the capital controversies. Unfortunately, this becomes essentially a onecommodity model, and is not, therefore, particularly enlightening. In any
case, the neo-Austrian model cannot incorporate a specific machine-goods
sector, or adequately treat fixed capital (Hagemann 1992), and so abstracts
from the main problems. This attempt by Hicks to analyse the traverse within
a neoclassical framework may be contrasted with the eorts of Kalecki and
Lowe, and illustrates the diculty of using the neoclassical approach to
meaningfully discuss disequilibrium phenomena (Lavoie and RamirezGaston 1977).
Lowe, in a return to the concerns of classical economics, sees the main
problem of economics as the description of the path of economic growth
(Lowe 1976), which is not normally of the steady-state equilibrium type.
He specifically analyses the traverse, and concentrates on the implications
of structural change. For this reason, he focuses on the nature of changes
in the structure of production and on intersectoral relations, in a manner
Traverse 357
reminiscent of the classical economists. To examine this problem, Lowe
developed a three-sector model which incorporated not only the concept of
historical time but also two important aspects of production; the specificity of capital goods and the importance of reproduction, which is necessary for the incorporation of intersectoral relations. Specificity is dealt with
by dierentiating two subsectors of the capital-goods sector. In the first
subsector, capital goods are produced which can either reproduce themselves or produce capital goods for the consumption-goods sector.
Although, at this stage, there is no distinction between the capital goods,
specificity becomes important when the capital goods produced in this
sector are installed, as on installation they lose their generality and, in an
irreversible process, become specific to the production of capital goods for
the consumption-goods sector. These capital goods may be considered a
separate branch of production. The capital-good output of this sector is
installed in the consumption-goods sector to produce consumption goods.
Lowe uses this model to examine the nature of the traverse. From an
initial stationary state, the implications for the traverse of changes and
restrictions on variables are analysed. The model is used to consider the
structural changes within the capital-goods sector which are necessary to
facilitate, for example, changes in technology and in the rate of growth of
the labour force, and their implications for intersectoral relations.
An important conclusion to emerge is Lowes demonstration that,
although there may very well be a traverse which leads to a new fullemployment steady state, it is unlikely to be achieved within a decentralized
market system. This, in part, results from the market transmitting the
wrong signals in terms of the optimal structure of production and intersectoral flows.
While the Hicks/Lowe traverse analysis concentrates on the supply-side
questions of the responses of the structure of production, Kaleckis
emphasis was on the demand side. For Kalecki, the structure of demand
was the key to the adjustment process, with the essence of the problem
being what happens to the composition of demand as a result of the
changes in the distribution of income during the cycle. As a result, although
Kalecki, like Lowe, disaggregated the economy into three sectors, the disaggregation served dierent purposes. For Kalecki the division of the
economy into a capital-goods sector and two consumption-goods sectors,
dierentiating workers consumption from that of capitalists, was the result
of his emphasis on the problems associated with realization in the form of
eective demand.
He did, however, share Lowes concern with reproduction and with intersectoral relations, but concentrated on flows of commodities and of
incomes between sectors. Kalecki saw the main determinant of income and
358
Traverse
growth in mature capitalist economies as being the level of eective

demand. Crucial to this was the dual role of investment, which, on the one
hand was part of eective demand (so that the higher the level of investment, the greater the level of employment in that period), while on the
other, it contributed to the creation of extra capacity (so that the higher the
level of investment, the larger would be the problem with achieving full
employment in subsequent periods). This paradox, according to Kalecki,
struck at the heart of the capitalist system: The tragedy of investment is
that it causes crisis because it is useful. Doubtless many people will consider
this theory paradoxical. But it is not the theory which is paradoxical, but
its subject the capitalist economy (Kalecki 199097, I, p. 318).
Although Kalecki concentrated on the role of eective demand in his
analysis of capitalist economies, in his work on socialist economies the
structure of production, rather than eective demand, was seen as the
important constraint on economic activity. Here he came much closer to
the traverse analysis of Hicks and Lowe, and in many ways their eorts are
complementary. In a capitalist economy, a reduction in investment causes
a reduction in profits which feeds through to a multiplied reduction in
income and aggregate demand. Kalecki contrasts this with the eects of a
reduction in investment in a socialist economy, where he argues there need
be no problem with eective demand:
The workers released from the production of investment goods would be
employed in the consumption goods industries. The increased supply of these
goods would be absorbed by means of a reduction in their prices. Since profits
of the socialist industries would be equal to investment, prices would have to be
reduced to the point where the decline in profits would be equal to the value of
the fall in investment. (Kalecki 199097, II, pp. 2545)
In his analysis of the structure of investment in socialist economies, he

acknowledged the possibility of short-run problems in adjustment caused
by capacity bottlenecks, in the sense of too much (or too little) capacity in
the capital-goods sector. It is here that Kaleckis work touches on issues
raised by Hicks and Lowe. Using a two-sector model, and dierentiating
investment in the capital-goods sector from aggregate investment, he
showed that changes in the growth rate of the economy will necessitate
deviations between the growth rate of investment and that of the economy,
during the transition period. However, there exists a ceiling to the deviation of the rate of growth of investment from that of national income
which is determined by the productive capacity of the investment sector
(Kalecki 199097, IV, p. 102).
Kaleckis two-sector model suers from its inability suciently to disaggregate the structure of the investment-goods sector. This is of particular
Treatise on Probability
359
importance when the problem is that of dierential growth rates between

the consumption- and investment-goods sectors, with investment goods
being provided to both sectors. It is here that Lowes model can supplement
Kaleckis discussion of structural problems, as well as showing the diculty of getting rid of excess capacity.
There are many other examples of traverse analysis in economic theory.
The literature on path-dependency, hysteresis, cumulative causation and
lock-in are some examples. The general conclusion of this literature is that,
without serious analysis of the traverse, all economic theory utilizing some
concept of equilibrium (including the long-run equilibrium analysis of the
Sraans) is vacuous. Without some demonstration that there are forces in
the economy which push it to equilibrium, without influencing the position
to which the economy is gravitating, it is dicult to foresee any useful role
for such equilibrium theory. However, such a demonstration is unlikely, as
without a visible hand, the invisible hand is likely to guide us on to the
wrong path; this is perhaps the most important conclusion from the analysis of the traverse (Halevi and Kriesler 1992, p. 233)
P K
See also:
Dynamics; Expectations; Growth Theory; Kaleckian Economics; Sraan Economics; Time
in Economic Theory.
References
Hagemann, H. (1992), Traverse analysis in a post-classical model, in J. Halevi, D. Laibman
and E. Nell (eds), Beyond the Steady State: A Revival of Growth Theory, Basingstoke:
Macmillan, pp. 23563.
Halevi, J. and P. Kriesler (1992), An introduction to the traverse in economic theory, in J.
Halevi, D. Laibman and E. Nell (eds), Beyond the Steady State: A Revival of Growth Theory,
Basingstoke: Macmillan, pp. 22534.
Hicks, J.R. (1965), Capital and Growth, Oxford: Oxford University Press.
Hicks, J.R. (1973), Capital and Time: a Neo-Austrian Theory, Oxford: Clarendon Press.
Kalecki, M. (199097), Collected Works of Michal- Kalecki, Oxford: Clarendon Press.
Lavoie, M. and P. Ramirez-Gaston (1997), Traverse in a two-sector Kaleckian model of
growth with target-return pricing, Manchester School, 65 (2), 14569.
Lowe, A. (1976), The Path of Economic Growth, Cambridge: Cambridge University Press.
Robinson, J. (1974), History versus equilibrium, in J. Robinson, Collected Economic Papers,
Vol. V, Oxford: Blackwell, 1979, pp. 4858.
Setterfield, M. (1995), Historical time and economic theory, Review of Political Economy, 7
(1), 127.
The Treatise on Probability (Keynes 1921) is J.M. Keyness main philosophical work. The book had a long gestation period. The key idea was
360
advanced in an undergraduate paper of 1904, before being expanded into

a dissertation, awarded in 1909, for a fellowship at Kings College,
Cambridge. Publication was envisaged shortly thereafter, but extensive
revisions and the First World War delayed its appearance until 1921. At the
time, it established Keynes as a leading authority in the philosophy of probability, the ongoing interest in the work being indicated by at least six
reprints (1929, 1943, 1948, 1952, 1957 and 1963). Its influence continues
nowadays, but in a weaker vein.
It is not a mathematical treatise on the probability calculus but a wideranging philosophical work which could easily have been entitled Treatise
on Reason or Treatise on Logic. In general terms, its significance is threefold. Firstly, it is a pioneering work advancing the earliest systematic exposition of the logical theory of probability, and its dierences from rival
theories (on which, see Weatherford 1982). Second, this theory of probability forms part of a broader theory of rational belief and action under
uncertainty. Third, Keyness philosophizing across a range of topics throws
light on various aspects of his thinking in economic theory, economic
policy, politics and the arts.
Keynes began his analysis of probability and rational belief with a question. How are we to understand the large class of arguments encountered
in many spheres of life that we regard as rational in some sense and yet we
know are non-conclusive (or non-deductive)? His answer was that rationality implies a connection with logic, and non-conclusiveness implies a connection with probability. The synthesis of these two ideas led to his
conception of probability as a general theory of logic, the subject matter of
which was rational but non-conclusive argument. Traditional or deductive
logic was embraced by the general theory as a special case.
Every argument, whether non-conclusive or conclusive, proceeds from a
set of premises, h, to a conclusion, a. Probability, in Keyness theory, is concerned with the relation between the two propositions, h and a. In the
general (non-conclusive) case, h lends some support to a but not complete
or conclusive support; that is, the premises partly entail the conclusion, but
are insucient for complete entailment. Keynes postulated that the relation
between h and a is a logical relation, which he called the probabilityrelation. It is a relation of partial support or entailment which, in his view,
belongs as much to logic as does the complete entailment of deductive
logic. The distinctive symbol Keynes adopted for probability is a/h, read as
the probability of a on premises h, this symbol emphasizing the data dependence of probability.
The connection between probability and rational belief was made as
follows. Assuming that h were true and that a deductively followed from h,
then the probability of a on h would be unity (a/h1) and it would be ratio-
361
nal to believe a with complete certainty. However, if h were true and only gave
partial support to a, the probability of a on h would be less than unity (a/h
1) and it would only be rational to believe a with a degree of certainty, this
degree of certainty being given by the probability. If, for example, it has been
raining for a week and no information indicates a cessation in the next few
days, the probability of the proposition, it will rain tomorrow, will be high,
and it will be rational to believe this proposition to a high degree, though not
with complete certainty since it cannot be deduced from the available information. Notice that rationality of belief in a is not tied to the truth of a. The
fact that a might later turn out to be false does not mean it was irrational, on
the available evidence, to believe a to some degree in the first place.
Keyness logical probability-relations express three aspects of arguments
degrees of partial inference (the extent to which a may be inferred from
h), degrees of rational belief (the extent to which it is rational to believe a,
given h), and degrees of certainty. The limits at either end of these degrees
are the probabilities of unity and zero, both of which are given by deductive logic. Unit probability corresponds to full entailment and complete
certainty. If a is fully entailed by h, then a/h1. Zero probability corresponds to contradiction or logical impossibility. If a is the contradictory
of a, then a/h0 means a/h1, that is, h fully entails the contradictory
of a, so that a is impossible and requires complete disbelief. Bounded by
these two extremes (0a/h1) is the densely populated universe of
Keyness probability-relations.
Such probabilities are always objective and never subjective. Their objectivity derives from their status as relations of logic. Between any pair of
propositions, a and h, the logical relation is unique and fixed independently
of personal opinion or psychological belief. Probabilities are thus members
of an immutable, non-natural realm of logical relations transcending
human subjectivity.
Knowledge of logical relations is arrived at by intellectual intuition, that
is, by careful reflection on the support h gives to a, or by mental insight into
the realm of logical relations. However, since mental ability varies across
individuals and history, not all logical relations will be known to all individuals at all times. Depending on our powers of logical insight, some probabilities will be known and some unknown.
The measurement of probabilities is an intriguing aspect of Keyness
theory. While other theories reduce probabilities to numerical (and hence
universally comparable) form, Keyness ordering of the probability space is
far more complex. Three types of comparative relations are postulated:
1.
Cardinal comparison, which generates the relatively minor class of

numerical probabilities. These only exist under the restrictive condition
362
2.
3.
of equiprobability, which is established by careful use of the principle
of indierence.
Ordinal comparison, which generates the much bigger class of nonnumerical probabilities. This class consists of many separate, incommensurable series whereby probabilities belonging to the same series
are comparable in terms of greater or lesser, but probabilities belonging to dierent series are generally incapable of being compared in
magnitude.
Non-comparability, which typically exists between numerical and nonnumerical probabilities, and between non-numerical probabilities in
dierent series.
Such heterogeneity in the ordering relations between probabilities greatly

restricts the scope of the probability calculus.
Keynes proposed that a second, independent variable, called the weight
of argument, is also relevant to rational belief under uncertainty. Weight
is positively associated with the data, h. It provides a measure of the wellfoundedness of non-conclusive argument and hence of the confidence that
may be placed in the argument. Arguments of high weight are based on
much information, are well founded and hence merit high degrees of confidence, regardless of whether their probability is high or low. Arguments
of low weight rest on little information, are poorly founded and deserve low
degrees of confidence even if their probability is high. All other things
being equal, rational agents should choose the argument with the greatest
weight.
After outlining the fundamental ideas in Part I of the book, Keynes
extended the discussion to associated issues in the remaining four parts. In
Part II, he undertook the necessary task of deriving the probability calculus. Using definitions and axioms based on probability as the relation, a/h,
he first arrived at the theorems of deductive logic, and then turned to probable inference to derive the theorems of the addition and multiplication of
probabilities and of inverse probability.
In Part III, Keynes observed that logicians had not oered satisfactory
accounts of the processes of induction that lay behind numerous arguments. Believing that every inductive argument was probabilistic in nature,
he sought to explore the analysis and logical justification of induction. His
procedure was, first, to dissect inductive methods; second, to isolate the
fundamental assumption(s) on which induction rested; and, third, to
inspect these assumption(s) for their truth or self-evidence without resorting to appeals to experience, which would involve circularity. The main
assumption he saw underpinning induction was the principle of limited
independent variety, which essentially says that the variety of the universe
363
is limited because it derives from a finite number of independent generators. He then cautiously suggested that this principle was self-evidently
true, thereby accepting the existence of synthetic a priori knowledge which,
along with the role of intuition in his theories of probability and ethics,
places the epistemology of his philosophical work in the rationalist camp.
In keeping with much philosophy of the time, Keyness treatment of
chance and randomness in Part IV is based on determinism. On this view,
chance, randomness or indeterminateness are not objective or inherent
characteristics of the universe, but are purely subjective phenomena deriving from human ignorance. It is our lack of knowledge of the relevant
causal order which leads us to say that certain events are due to chance or
are random occurrences, when actually they are the product of causal influences of which we are currently ignorant.
Also in Part IV, Keynes addressed the vital question of what ought we
to do?, but in only one chapter. His answer, which combined his logical
theory of probability with the ethics of G.E. Moore, may be described as a
probabilistic form of consequentialism. On this view, rational or right
action is that which is judged to produce the greatest amount of probable
goodness on the whole, appropriate attention being paid to weight and
moral risk. Broadly speaking, the assessment of an action involves a listing
of its possible consequences and a means of combining the probabilities
and values of these consequences into probable goodness on the whole.
When numerical magnitudes can be ascribed to probabilities and values the
procedure is straightforward, but, when probabilities and/or values are
non-numerical in nature, Keynes appealed to direct judgement or intuition
as the means by which we arrive at the probable goodness of the whole. If
such judgement was not within our power, he accepted the matter as indeterminate. In such cases, suggestions were made that reason can turn to
second-best options such as conventions, rules and caprice.
The concept of moral risk directs attention to the probability of failure
or non-occurrence. It is most relevant to extreme cases where probable
values derive from a combination of very low probabilities and very large
gains. Even if the probable value is acceptable, the probability of failure is
very high, as is the probability of losing the money or eort wagered on the
outcome. If this loss would have catastrophic eects, rational individuals
should be prudent and pursue actions with lower risks of loss. One should
not bet all ones wealth on a horse race; winning might produce enormous
gain, but losing would result in pauperism.
In the final part, Keynes investigated the logical basis of statistical inference. He analysed the Law of Large Numbers, the theorems of Bernoulli,
Poisson and Chebyshev, Laplaces rule of succession, and the methods of
Lexis, before concluding, all too briefly, with an outline of a constructive
364
theory. His object was to uncover and discredit invalid inferences and to
clarify and support valid inferences, his guiding principle throughout being
that logic was the master and mathematics the servant. What drew his
censure, here and in relation to mathematical methods generally, was loose
thinking and inappropriate application, not valid and logically consistent
use.
Questions naturally arise as to the existence of connections between
Keyness philosophical thought and his non-philosophical thought in areas
such as economics, politics and the arts. Of central importance is the relationship between his major philosophical work (the Treatise on Probability,
hereafter the Treatise) and his major economic work, The General Theory
of Employment, Interest and Money. Some of the key issues discussed have
been whether substantial connections exist between: (i) the treatment of
probability and uncertainty in the Treatise and the treatment of uncertainty and expectations in the General Theory; (ii) the non-neoclassical
rationality of the Treatise and the non-neoclassical treatment of agent
behaviour in the General Theory; (iii) the epistemology of the Treatise and
his epistemological remarks in the General Theory and elsewhere; (iv) the
approach taken to quantitative matters and formalism in the Treatise and
his comments on mathematical methods in economics and econometrics in
other writings; and (v) the ethics and theory of right action of the Treatise,
and his thoughts on political philosophy and party politics in other works.
Such questions have generated considerable controversy. Some writers
argue that various strong threads of continuity exist between the conceptual framework of the Treatise and major themes in Keyness other
thought. Others claim that his philosophical thinking changed significantly
after 1921, thereby creating serious discontinuities between the Treatise and
the General Theory. Debate has arisen over whether he remained an adherent of the logical theory of probability, or was converted shortly afterwards
to Frank Ramseys subjectivist theory. Discussion also surrounds rationality, irrationality and Freudian influences in his thought. For two views, out
of the many available, on the relations between Keyness philosophy and his
other writings, see Carabelli (1988) and ODonnell (1989).
R OD
See also:
Agency; Econometrics; Expectations; Keyness General Theory; Uncertainty.
References
Carabelli, A.M. (1988), On Keyness Method, London: Macmillan.
Keynes, J.M. (1921), A Treatise on Probability, London: Macmillan. Republished in 1973
(with dierent pagination) as Volume VIII of The Collected Writings of John Maynard
Keynes, London: Macmillan.
365
ODonnell, R.M. (1989), Keynes: Philosophy, Economics and Politics. The Philosophical
Foundations of Keyness Thought and their Influence on his Economics and Politics, London:
Macmillan.
Weatherford, R. (1982), Philosophical Foundations of Probability Theory, London: Routledge
& Kegan Paul.
Uncertainty
Economics texts commonly define a situation of risk as one in which an
individual with a decision to make is able to assign numerical probabilities
to all outcomes that could possibly follow from that decision. If probabilities cannot be assigned, a situation of uncertainty is then said to obtain.
These texts characteristically focus their attentions on the risk situation
thus specified: having defined uncertainty, they simply ignore it. Some
mainstream theorists take a dierent tack and assume that economic agents
can always identify, and assign probabilities to, the whole range of possible
future states of the world. This group has no need to make a distinction
between risk and uncertainty: for them, uncertainty itself is inherently
quantifiable.
The passion for numerical probability shown by mainstream economists
is no accident. For their central concept of rational optimizing choice,
developed under assumptions of perfect knowledge, can seem somewhat
irrelevant to the real world, where certainty is never obtainable. However,
if knowledge of probabilities attaching to alternative possible future scenarios is assumed to be available, then (statistically) expected values can be
calculated. These can, in turn, provide a basis for a theory of optimal decisions in a world in which perfect knowledge is lacking. In this way, many
orthodox economists believe they have remedied a major perceived deficiency in their theoretical apparatus.
For Post Keynesian economists, in contrast, the future is characterized
by fundamental uncertainty. They take their cue from Keynes, who rejected
the possibility of a theory of optimal choice under uncertainty, observing
that human decisions aecting the future . . . cannot depend on strict mathematical expectation, since the basis for making such calculations does not
exist (Keynes 1936, pp. 1623). Drawing primarily on the work of Paul
Davidson, the discussion which follows tries to explain the concept of fundamental uncertainty in broad terms and to show why it occupies such an
important place in Post Keynesian economic theory. As a first step, it is
appropriate to review some important concepts developed by George
Shackle, since his theories have been a major influence on the way Post
Keynesian economists in general think about uncertainty and on
Davidsons ideas on the subject in particular.
Any question about the future represents what Shackle (1968 [1955],
p. 63) calls an experiment. He defines a divisible experiment as one consisting of a series of trials all in some sense alike and each important only as
366
Uncertainty
367
it contributes to the total result of the series (p. 64). For instance, a firm
might regard any sale to a prospective customer as a trial. Its sales experiment would then be a divisible one, if (as is likely) its management were
primarily interested in the total revenue generated by its sales eort and not
in the outcome of its dealings with any specific customer.
Shackle accepts that it might in principle be possible to calculate numerical probabilities in divisible experiments, provided that frequency distributions could be derived from suitably uniform circumstances and a
large number of [past] repetitions (p. 4). His point, however, is to challenge what that would achieve. For, suppose our firm could draw up reliable
frequency data from individual customer records and so meet these preconditions for calculating numerical probabilities. Its total monthly sales
revenue could then be forecast with considerable accuracy and could not be
described as being a matter over which its management was in any real
sense uncertain. Thus, Shackle argues, a situation in which numerical probabilities could be calculated has nothing whatever to do with ignorance or
uncertainty: it is knowledge (p. 4). His argument implies that the orthodox
economists whose approach was outlined above delude themselves in imagining they have developed theories that have anything to do with real
uncertainty.
An even more potent concept developed by Shackle is that of the crucial
experiment. This he explains as follows:
An experiment can be such that . . . the making of it will radically alter the situation . . . so that it will subsequently be impossible . . . to perform another
experiment of a relevantly similar kind. Napoleon could not repeat the battle of
Waterloo a hundred times in the hope that, in a certain proportion of cases, the
Prussians would arrive too late. His decision to fight on the field of Waterloo was
what I call a crucial experiment . . . Had he won, repetition would . . . have been
unnecessary; when he lost, repetition was impossible. (Shackle 1968 [1955], p. 25)
Shackles main point is not that the average outcome of a large number of
similar experiments is hypothetical when the experiment is crucial. Rather,
it is that any such average is of no interest to the crucial experimenter,
because the latter is never going to be in a position to oset a poor outcome
now with better ones later. Shackles position can perhaps be summarized
as follows: when numerical probability calculations are relevant, the situation is not uncertain; when the situation is uncertain, they are not relevant.
Shackle insisted that The diculty of obtaining reliable frequency ratios
is the lesser strand of my argument (1968, p. 4). Although drawing on
Shackles concepts, Davidsons eorts have, however, been directed at
showing just how serious that diculty is. Now Keynes (1973 [1921], p.
468) castigated the Professors of probability (who are) often and justly
368
Uncertainty
derided for arguing as if nature were an urn containing black and white
balls in fixed proportions. Nevertheless, Davidson insists that the conceptions of these professors modern-day counterparts must be used if the case
against them is to be demonstrated. To follow Davidsons analysis it is
therefore helpful to begin by considering what the world would be like if
nature were akin to Keyness urn.
The outcome of any drawing from the urn will be a stochastic (random)
event, so that we can possibly think of a series of outcomes in a succession
of drawings as resulting from a stochastic process. Furthermore, if these
drawings take place at consecutive points in time, they will yield a time
series, from which an average outcome, a time average, can be calculated.
This series could be infinite. It is not so easy to conceive of all possible outcomes of a drawing at a single point in time as existing simultaneously in a
universe or ensemble (Davidson 198283, p. 189) of alternative logical
worlds. However, having made this conceptual leap, we can grasp the idea
of a space average, the mean value taken by a (finite or infinite) set of such
drawings (ibid.).
On this basis, we can begin to understand the following definition of an
ergodic process, Davidsons formal equivalent to natures urn: If the stochastic process is ergodic, then for an infinite realization [that is, set of drawings from the urn], the time and space averages will coincide. For finite
realizations . . . the space and time averages tend to converge (Davidson
1994, p. 90; see also Davidson 198283, p. 185). The importance of this
concept in Davidsons work is measured by the frequency with which more
or less this same terse definition is restated in his writings. It is the basis
on which he asserts the existence of what he calls the ergodic axiom,
the assumption that economic events are broadly governed by ergodic
processes. This axiom, he suggests, is the ultimate logical foundation of
orthodox theorizing (Davidson 1994, pp. 8990; 1996, p. 494). For the convergence of averages in an ergodic system implies that, over a long enough
period of time, events will tend to follow repetitive patterns, and it is central
to Davidsons argument that just this kind of conception lies behind the
orthodox belief in an inherent tendency in market economies towards a
long-run (full-employment) equilibrium (Davidson 1996, p. 496).
Of course, Davidsons purpose in highlighting the importance to orthodox theory of an ergodic axiom is only to demonstrate that belief in it is
untenable. This he seeks to do on both empirical and logical grounds. Now,
simply stated, a time series is stationary if all observations that go to make
it up are drawn from distributions with the same mean and variance, and
his empirical objection is that real-world time-series data often fail to
exhibit evidence of stationarity. As he points out, non-stationarity is, formally, a sucient condition for non-ergodicity (ibid., pp. 4945).
Uncertainty
369
Davidsons logical objection is perhaps the more fundamental. He

observes, referring to Shackles concept, that When agents make crucial
decisions, they necessarily destroy any ergodic processes that may have hitherto existed (Davidson 198283, p. 192). Furthermore, if crucial decisions
are all pervasive (Shackle 1968 [1955], p. 63) or even just very common
(Davidson 198283, p. 192), then the economy as a whole will be in a state
of constant change as economic agents continually alter the pre-existing
state of things by taking them. That will in turn represent a sucient condition for a non-ergodic world (ibid.; see also Davidson 1996, pp. 497, 500).
In that world, decision makers will typically operate in an environment of
fundamental uncertainty, since it will be impossible to base a whole range
of significant economic decisions on forecasts derived from past data.
The significance of this analysis for Post Keynesian theory is immense.
For, as Davidson (1994, pp. 934; 1996, pp. 501, 506) insists:
1.
2.
3.
4.
individuals will exhibit long-term liquidity preference only under conditions of fundamental uncertainty;
the Post Keynesian contention that money is not neutral even in the
long run assumes the possibility of long-term liquidity preference;
the Post Keynesian critique of the notion that market economies have
an automatic tendency towards full-employment equilibrium is based
on a denial of the neutrality of money; and
Post Keynesian policy analysis, with its emphasis on government intervention and the need for protective institutions, arises out of that critique.
As has been shown, the essential premise of the first of this arguments
four steps is the proposition that the crucial experiment is, at the very least,
a common phenomenon in the economic sphere. But if crucialness is
important in the real world, why do orthodox theorists fail even to acknowledge it as a possibility? The answer would seem to be that, conceiving nature
as an urn like Keyness professors of probability, they cannot do so. For that
would be tantamount to accepting that the proportions of black and white
balls in the urn could be changed by human decisions and would make it
impossible to view cause and eect simply in terms of the operation of stochastic processes.
The idea of a stochastic process, it is reasonable to argue, implies a metaphysical view of the world that is at odds with human experience. This is
because:
the existence of a process implies that the outcome today stands at the end, in
real time, of a chain of causation and may be dierent from what it was yesterday because of prior changes somewhere along the length of that causal chain.
370
Underconsumption
Processes are in reality the polar opposite of repeated drawings from some eternally subsisting distribution. In other words, in place of a genuine process there
is, within the stochastic scheme of things, only the isolated, accidental outcome,
dissociated from any determining cause/eect relationship with anything that
has gone before. (Glickman 199798, p. 262)
Davidson has noted that If observed economic events are not the result of
stochastic processes then objective probability structures do not even fleetingly exist (1994, p. 90). The historical/causal conception of a process just
outlined oers a realistic alternative account of how economic events are
determined that can help us understand why uncertainty is fundamental in
economics.
M G
See also:
Agency; Expectations; Keyness General Theory; Liquidity Preference; Non-ergodicity;
Treatise on Probability.
References
Davidson, P. (198283), Rational expectations: a fallacious foundation for studying crucial
decision-making processes, Journal of Post Keynesian Economics, 5 (2), 18297.
Davidson, P. (1994), Post Keynesian Macroeconomic Theory, Aldershot: Edward Elgar.
(4), 479508.
Glickman, M. (199798), A Post Keynesian refutation of ModiglianiMiller on capital structure, Journal of Post Keynesian Economics, 20 (2), 25174.
Macmillan.
Keynes, J.M. (1973 [1921]), A Treatise on Probability, reprinted as The Collected Writings of
John Maynard Keynes, Vol. VIII, edited by D.E. Moggridge, London: Macmillan for the
Shackle, G.L.S. (1968 [1955]), Uncertainty in Economics, Cambridge: Cambridge University
Press.
Underconsumption
Although Keynes wrote about underconsumption at some length, the term
is seldom used these days by Post Keynesians. Some of the central issues
continue to resonate, however, in discussions of economic growth and the
relationship between wages and employment.
Keynes devoted chapter 23 of the General Theory to Notes on Mercantilism, the usury laws, stamped money and theories of under-consumption,
revealing considerable sympathy for seventeenth- and eighteenth-century
work on the dangers of excessive saving. He also quoted Thomas Malthus
and provided a critical commentary on such later underconsumptionists as
Underconsumption
371
J.A. Hobson and Major C.H. Douglas. The primary evil identified by
these writers, Keynes concluded, was a propensity to save in conditions of
full employment more than the equivalent of the capital which is required,
thus preventing full employment except when there is a mistake of foresight (Keynes 1936, pp. 3678). By the capital that is required he seems to
have meant the addition to the capital stock needed to cater for the increase
in output as a whole (not just consumption goods), while the mistake in
foresight referred to the belief of Hobson and other underconsumptionists that unemployment would normally result unless entrepreneurs
wrongly invested more than was justified by this requirement.
In chapter 22, Notes on the trade cycle, Keynes declared his support for
all sorts of policies for increasing the propensity to consume (ibid., p. 325),
and oered a numerical example to illustrate what might be required:
[If] the average level of output of to-day is 15 per cent. below what it would be
with continuous full employment, and if 10 per cent. of this output represents
net investment and 90 per cent. of it consumption if, furthermore, net investment would have to rise 50 per cent. in order to secure full employment, with the
existing propensity to consume, so that with full employment output would rise
from 100 to 115, consumption from 90 to 100 and net investment from 10 to 15:
then we might aim, perhaps, at so modifying the propensity to consume that
with full employment consumption would rise from 90 to 103 and net investment
from 10 to 12. (ibid., pp. 3256)
Quite how this was to be achieved, Keynes did not say. The earliest underconsumptionists had called for a larger share of national income to go to
the idle rich, who could be relied upon not to save very much of it.
Subsequently, liberal and socialist writers urged redistribution in favour of
the poor, on the grounds that the propensity to save out of wages and salaries was considerably higher than that out of profits (and rents).
This is where Keynes or at least his left-wing supporters met Marx.
The brief references to Marxism in the General Theory are disparaging,
though in a 1933 draft Keynes had been much less critical. There was in fact
a long tradition of underconsumptionism in the Marxian literature on economic crises and not surprisingly it featured prominently in analyses
of the Great Depression by socialist theorists like Otto Bauer, Natalie
Moszkowska and Eugen Varga. The inexorable tendency of capitalist production, they argued, was towards an increase in the rate of exploitation,
and hence in the share of profits in net output. This had become even more
pronounced in the latest, monopoly stage of capitalism, which was dominated by what Paul Baran and Paul Sweezy termed the law of the rising
surplus. The decline in the wage share had rendered the system liable to
crises of overaccumulation, with the capital stock growing more rapidly
than was warranted by the increase in consumption. Thus, even if higher
372
Underconsumption
saving were to stimulate higher investment, the situation was inherently

unsustainable. Any boom would end in a severe cyclical downturn, or
perhaps (in the most dramatic versions of the argument) in the complete
breakdown of the system. The maturity of capitalism was therefore synonymous with stagnation, according to the Austrian left Keynesian Josef
Steindl (King 2002, chapter 2).
Of all Keyness close allies, it was Joan Robinson who was most receptive to these ideas. While she criticized Marx for trying to work out a theory
of crisis in which Says Law continued to hold, she also saw considerable
merit in his reproduction models. They could be used, Robinson suggested,
as the basis for a theory in which
consumption by the workers is limited by their poverty, while consumption by
the capitalists is limited by the greed for capital which causes them to accumulate wealth rather than to enjoy luxury. The demand for consumption goods (the
product of group II) is thus restricted. But if the output of the consumptiongoods industries is limited by the market, the demand for capital goods (group
I) is in turn restricted, for the constant capital of the consumption-good industries will not expand fast enough to absorb the potential output of the capitalgood industries. Thus the distribution of income, between wages and surplus, is
such as to set up a chronic tendency for a lack of balance between the two groups
of industries. (Robinson 1942, p. 49)
In this interpretation Robinson was heavily influenced by Michal- Kalecki

who, though not himself an underconsumptionist, had absorbed the
closely related work of Mikhail Tugan-Baranovsky and Rosa Luxemburg
while still a young man in Poland. In a celebrated passage Kalecki wrote of
the tragedy of investment, which adds both to eective demand and to the
capital stock, simultaneously stimulating growth and undermining it
(Kalecki 1939 [1991], p. 284).
In formal terms, there are certain similarities between models of underconsumption and the accelerator principle that is central to HarrodDomar
growth theory, itself a major influence on Post Keynesian thinking about
economic growth. It is significant that both Roy Harrod and Evsey Domar
acknowledged their debt to Hobson. Keynes analysed what happened when
savings (of the preceding period) are not invested, Domar wrote. Hobson,
on the other hand, went a step further and stated the problem in this form:
suppose savings are invested. Will the new plants be able to dispose of their
products? Such a statement of the problem was not, as Keynes thought, a
mistake. It was a statement of a dierent, and possibly also a deeper
problem (Domar 1957, p. 52). Some intricate analytical issues are involved
here, in particular the precise specification of the accelerator principle (does
investment depend on the rate of increase in consumption or the rate of
increase in total output?), and the relationship between underconsumption
Underconsumption
373
and what has been termed the underinvestment theory that is implicit in
the HarrodDomar model (Schneider 1996, pp. 7783).
If growth theory is one area in which underconsumptionist thinking continues to influence Post Keynesian macroeconomics, the other is wage and
employment theory. Sidney Weintraub (1956) approached this question
from the perspective of Keyness aggregate supplyaggregate demand analysis. An increase in the general level of money wages would shift the aggregate supply curve upwards, by increasing firms costs of production, and
would therefore tend to reduce employment. But it might also shift the
aggregate demand curve up, through increasing consumption by wage
earners; alternatively, it could reduce aggregate demand by depressing the
marginal eciency of capital and discouraging investment. Weintraub distinguished three possibilities: the classical case, where employment falls as
money wages rise; what he termed the Keynesian case, where shifts in aggregate supply and aggregate demand cancel each other out; and the underconsumptionist case, where employment increases as money wages rise.
While Weintraub focused on money wages, Amit Bhaduri and Stephen
Marglin (1990) provided a formal analysis of the eects on aggregate demand
of an exogenous change in the real wage. They distinguished stagnationist
and exhilarationist regimes, the former (which corresponds to the classic
underconsumptionist position) prevailing when investment responds weakly
to a reduction in profitability while consumption increases substantially if
real wages rise. The case for the stagnationists, they argued, is stronger in the
context of a closed economy than in an open economy, where eective
demand can be increased by a currency devaluation that lowers the real wage.
The policy issues that are involved here are evidently very important, and
potentially very divisive. Kalecki angrily criticized those self-proclaimed
workers friends who urged trade unions to accept money-wage reductions
in the interests of increased employment (Kalecki 1939 [1991], p. 318),
although this did not make him an advocate of wage inflation as the route
to full employment. Instead he suggested that higher money wages were
likely to generate price inflation, leaving real wages and real consumption
unchanged. Only a reduction in the product market and labour market
power of the capitalists, Kalecki concluded, would permanently raise real
wages and reduce the share of profits in national income.
Is it possible that both workers and capitalists might gain from an
increase in real wages that increases eective demand, thereby raising
capacity utilization and increasing total profits? Bhaduri and Marglin contrast the social democratic ideology of economic cooperation between the
classes with the conservative (and Marxian!) view that conflict over income
distribution is inescapable in any capitalist economy. In their model the
values of the parameters determine precisely when cooperation will give
374
Unemployment
way to conflict, and it proves impossible to establish an unambiguous association between the stagnationist regime and class cooperation on the one
hand, and the exhilarationist regime and class conflict on the other
(Bhaduri and Marglin 1990, Figure 3, p. 389).
The evidence on the relationship between real wages and aggregate
employment is also rather mixed. In the late 1930s Keynes was convinced
by friendly critics, among them Kalecki, John Dunlop and Lorie Tarshis,
that increased employment was associated with higher, not lower, real
wages. More recent research suggests that he (and they) may well have been
mistaken. Lavoie (199697) defends Kalecki, arguing that an increase in the
base real wage paid to fixed (or core) employees does indeed generate
greater employment. But the average real wage falls due to a rise in the
proportion of low-paid peripheral (temporary or variable) workers.
Empirically no less than theoretically, underconsumption remains one of
the more contentious questions in Post Keynesian economics.
J.E. K
See also:
Growth and Income Distribution; Growth Theory; Income Distribution; Kaleckian
Economics; Saving; Wages and Labour Markets.
References
Bhaduri, A. and S. Marglin (1990), Unemployment and the real wage: the economic basis for
contesting political ideologies, Cambridge Journal of Economics, 14 (4), 37593.
Domar, E.S. (1957), Expansion and employment, American Economic Review, 47 (1), 3455.
Kalecki, M. (1939 [1991]), Essays in the Theory of Economic Fluctuations, London: Allen &
Unwin; reprinted in Collected Works of Michal- Kalecki, Vol. I, Oxford: Clarendon Press.
Macmillan.
King, J.E. (2002), A History of Post Keynesian Economics Since 1936, Cheltenham, UK and
Lavoie, M. (199697), Real wages, employment structure, and the aggregate demand curve in
a Post Keynesian short-run model, Journal of Post Keynesian Economics, 19 (2), 27588.
Robinson, J. (1942), An Essay on Marxian Economics, London: Macmillan.
Schneider, M.P. (1996), J.A. Hobson, London: Macmillan.
Weintraub, S. (1956), A macroeconomic approach to the theory of wages, American
Unemployment
In capitalist economies, individuals and families are largely responsible for
providing for their own well-being. In all industrialized and many developing economies, most workers do not have the means of production to provide
for their own subsistence, but rather must obtain the means of purchase and
means of payment (money) necessary for buying the means of subsistence
Unemployment 375
by selling their labour power in the market. In addition, the requirement that
taxes be paid in government currency means that even those possessing the
means of production to provide for their own subsistence nevertheless must
usually enter the labour market to obtain that which is necessary to settle
their tax obligations.
Unemployment, the failure to obtain employment that earns wages or
salaries paid in money, thus has a dire impact on the jobless, and is also
associated with tremendous social and economic costs for society as a
whole. Whereas, in neoclassical economics, market systems possess an
inherent tendency to full employment, in Post Keynesian economics unemployment is seen as a normal feature of capitalist economies. The eective
demand problem means that capitalist economies have trouble attaining full
employment, while the structural change problem means that capitalist
economies have trouble maintaining full employment, even if it could be
attained. In addition, some Post Keynesians echoing Marx have identified the functionality of unemployment, which presents obstacles to Post
Keynesian economic policies to eliminate unemployment. Nevertheless,
Post Keynesian economics does suggest policies that might assist capitalist
economies in attaining and maintaining full employment, without resulting in other macroeconomic problems, such as inflation.
Unemployment has tremendous social and economic costs (see, for
example, Piachaud 1997). Unemployment causes permanent losses of
output of goods and services. The unemployed are faced with financial
insecurity, resulting in poverty and indebtedness. Certain kinds of criminal
activity are directly related to unemployment. Many studies have linked
unemployment to family disruption, suicide, ill health (physical and
mental), drug addiction, homelessness, malnutrition, poor prenatal care,
school dropouts, racial and ethnic antagonism, and other social problems
(see, for example, Jahoda 1982). Unemployment also dierentially aects
certain sectors of the population, so that disadvantaged minorities, those
with little education, and youth, for example, can suer from rates of unemployment two to ten times the overall rate.
Unemployment can also destabilize business expectations, as fears of
low demand cool private investment. Related to this, unemployment can
also lead to technological stagnation. If, as Marx and others suggest, high
levels of employment stimulate technical innovation, unemployment would
be associated with less innovation. Firms with high and stable levels of
demand have the resources and the incentive to support high-tech development; with high unemployment and thus cheap labour, firms lack the
resources and the incentive to retool. It has also been shown that unemployment leads to deterioration in labour skills. All of this suggests that unemployment may lead to lower productivity growth.
376
Unemployment
Unemployment is the direct and indirect cause of many social and economic problems. It can also lead to political instability. Since Keynes, Post
Keynesians have dedicated significant attention to the problems of unemployment. Since unemployment is the cause of so much social and human
misery, it is of great interest whether capitalist economies tend to full
employment or whether unemployment is a normal feature of capitalism,
and thus a target for government intervention.
In neoclassical economics, market systems tend to utilize all resources
fully, including labour. Perfectly flexible wages, prices and interest rates
constitute the self-adjusting mechanism that will tend to eliminate unemployed resources in the long run. In the neoclassical version of Says Law,
if there is unemployment wages will adjust to increase labour demand, and
interest rates will adjust to ensure that the excess of aggregate income over
aggregate consumption at the full-employment level of output will be
invested. There is no involuntary unemployment in the long run, unless
there are market imperfections such as sticky wages, government interference, or other institutional rigidities (for example, unions). For neoclassical economics, if there is unemployment, government should stay out and
let the market correct itself; if there are market imperfections, government
may promote conditions under which the self-adjusting mechanism works
most smoothly, for example, deregulation, anti-trust and so on.
In The General Theory of Employment, Interest and Money, Keynes overthrew Says Law and demonstrated the possibility and the likelihood that
market systems do not tend to fully utilize resources, even under competitive conditions, due to insucient eective demand. Keynes criticized the
neoclassical theory of saving and investment, arguing that traditional loanable funds theory holds income constant when looking at savings and
abstracts from expectations when analysing investment. If aggregate saving
is primarily a function of income, not interest rates, and investment is determined by the expected profitability of investors and lending institutions,
then saving does not determine investment through variations in the rate of
interest, and the economy does not automatically tend to full employment.
Instead private investment determines savings through changes in income,
but there is no reason to expect that the full-employment level of investment
will always be undertaken. Keyness analysis takes place in historical rather
than notional (or logical time). The past is unchangeable and the future is
unknown and unknowable. Money must be understood as an institution for
dealing with radical uncertainty. The result is that capitalist economies tend
to operate with excess capacity and unemployment. It is therefore unlikely
for a capitalist economy, on its own, to attain full employment.
But capitalist economies have problems maintaining full employment,
even if it could be attained, due to ongoing structural and technological
Unemployment 377
change, such as changes in labour supply and the supply of natural
resources, labour- and capital-displacing technical change, and changes in
the composition of final demand. An economy running at full capacity and
full employment would be unable to respond to such changes, and sectoral
imbalance is here added to aggregate imbalance as a further cause of unemployment. Bottlenecks and rigidities mean that full employment is likely to
be inflationary. Structural change will soon result in unemployment, as
technology displaces workers in one sector and fails to absorb them in
another, the formation of real capital fails to keep up with the pace of a
growing labour supply, or declining demand in one sector fails to be oset
by demand for new products. Works such as Pasinetti (1981) and Lowe
(1976) oer structural models that demonstrate the great unlikelihood of
capitalist economies maintaining full employment, even if it could be
attained.
The eective demand and the structural change problems are economic
causes of unemployment. But Post Keynesians such as Michal- Kalecki
have noted that there may also be political obstacles to full employment.
Since unemployment in Keynes is a negative byproduct of capitalism, it is
viewed as serving no purpose in the capitalist system and so is clearly undesirable for all. Kalecki, Josef Steindl and others, however, have highlighted
that unemployment may be functional in capitalism, an insight that is
drawn from Marxs analysis of the reserve army of labour.
In Marx, unemployment serves several functions. First, it provides the
system with a pool of available labour from which to draw when the pace
of accumulation increases. Second, unemployment serves to discipline
workers, who may not fear being laid o in an environment of full employment. Third, unemployment holds down wages, since one of the ways in
which unemployment disciplines workers is to decrease their bargaining
power and thus keep wages from rising. Thus, in this view, unemployment
is not only a natural byproduct of capitalism, it is essential to its smooth
operation.
Marx postulated a number of dierent components of the reserve army
of labour. The latent reserve includes those currently outside of the
market system, either performing unpaid household labour or eking out a
meagre subsistence in the periphery of Third World economies. The stagnant reserve includes those who are almost never employed, boom or bust.
Members of the floating reserve alternate between employment and
unemployment, with the ups and downs of the business cycle. Paupers is
the term Marx used to identify those who are now often referred to as the
underclass. Recently, it has been suggested that changes in global capitalism have rendered some of these components no longer functional. This
has resulted in an environment conducive to policies that may promote the
378
Unemployment
elimination of the emerging surplus population with scary genocidal and

racist connotations (Darity 1999).
Policies to address unemployment must recognize both the eective
demand and the structural change problems, as well as the functionality of
unemployment and the emergence of a hard-core, unemployable sector no
longer functioning as a reserve army. Traditional Keynesian policies initially
attempted to stimulate aggregate demand through fiscal and monetary policies. Stimulating the private sector to full employment may address the
aggregate demand problem but not the structural change problem. In fact,
since the structural change problem emerges most forcefully at higher levels
of capacity utilization and employment, stimulating private sector demand
may increase the structural change problem. Some Post Keynesians would
utilize incomes policies to deal with some of the symptoms. Other routes
would include promoting public works and the socialization of investment.
These latter approaches, if designed correctly, may be more eective than
conventional fiscal stimulus in dealing with the structural change problem.
In the framework of a capitalist economy, full employment requires a policy
or a set of policies that can increase eective demand without bringing
on structural rigidity and that can eliminate unemployment while finding
some institutional mechanism for dealing with the functionality question.
Recently Post Keynesians have suggested that such a policy is available
in the form of a kind of permanent Works Progress Administration (Wray
1998). Hyman Minsky (1986) referred to this as government as employer
of last resort. Under such a policy, the government would provide a public
service employment (PSE) job to anyone ready and willing to work. As the
economy expands (contracts), the private sector demand for labour would
increase (decrease), and the PSE sector would shrink (grow). PSE workers
would be employed in all kinds of social and public services that would
benefit the community. Elimination of long periods of unemployment
would preserve and potentially enhance labour productivity. The social and
economic costs of unemployment due to income insecurity and poverty
would decline, and society would experience a significant benefit in the
form of less crime and other social problems associated with unemployment. The eective demand problem would be solved by maintaining
aggregate income at high levels, but the PSE approach, unlike traditional
demand stimulus, would address the structural change problem as well
(Forstater 1998). Instead of workers alternating between employment and
unemployment, sectoral and aggregate change would only alter the proportion of private and public sector employment. PSE also can address environmental problems. Stimulating the private sector to full employment
would surely result in greater pollution and exhaustible resource utilization,
while PSE activities may be designed to pollute less and use less fossil fuel.
Unemployment 379
Unemployment is at the root of many of the economic and social problems of capitalism. Some would argue that, instead of tinkering with capitalism, a new economic system should be sought in which the right to a job as
put forward in the United Nations Universal Declaration of Human Rights
is realized. Perhaps in a post-capitalist society the employmentmoney link
will be severed and a new mode of social and economic organization will
make unemployment extinct and irrelevant. Until such a time, however, there
is no reason not to go immediately to full employment with a guaranteed
public service job for anyone ready and willing to work.
M F
See also:
Economic Policy; Eective Demand; Employment; Full Employment; Kaleckian Economics;
Keyness General Theory; Says Law; Wages and Labour Markets.
References
Darity, Jr., William A. (1999), Who loses from unemployment?, Journal of Economic Issues,
33 (2), 4916.
Forstater, Mathew (1998), Flexible full employment: structural implications of discretionary
public sector employment, Journal of Economic Issues, 32 (2), 55763.
Jahoda, Marie (1982), Employment and Unemployment: A SocialPsychological Analysis,
Lowe, Adolph (1976), The Path of Economic Growth, Cambridge: Cambridge University
Press.
Minsky, Hyman P. (1986), Stabilizing an Unstable Economy, New Haven: Yale University
Press.
Pasinetti, Luigi (1981), Structural Change and Economic Growth, Cambridge: Cambridge
University Press.
Piachaud, David (1997), A price worth paying? The costs of unemployment, in J. Philpott
(ed.), Working for Full Employment, London: Routledge, pp. 4962.
Wray, L. Randall (1998), Understanding Modern Money: The Key to Full Employment and
Price Stability, Cheltenham, UK and Northampton, MA, USA: Edward Elgar.
Wages and Labour Markets

To most mainstream economists, the market for labour services is comparable to that for any other commodity. In accordance with individualistic
methodology, neoclassical economists posit many buyers (demand) and
sellers (supply), and wages are the price signals that pull together these economic agents within a labour market setting whose two presumed functions
are allocation and clearance. As long as there are no obstacles to individual
agents hedonistic pursuit of gain, wages fulfil their primary allocative function across markets by directing labour services to their most productive use
with the establishment of appropriate compensating dierentials. In the
absence of institutional constraints, wages are assumed to perform well
their crucial market-clearing function by means of price adjustment. In particular, movements in real wages would guarantee that the aggregate labour
market comes to rest either at full employment or at some natural level of
unemployment whose magnitude is believed to depend on the degree of
market imperfections that prevent labour market flexibility. The orthodox
conclusion, usually in the form of policy prescriptions, is that, to preserve
its self-correcting properties, the labour market must be designed by the
state in such a way as to eliminate all institutional features that hinder
market clearance and, more precisely, limit downward wage adjustment,
such as trade unions, minimum wages and transfers to the unemployed.
While some neoclassical economists, especially of the New Keynesian
variety, have come to recognize that this peculiar conception of the labour
market is somewhat problematic and have pointed to the latters unique
characteristic as a social institution with features quite distinct from the
market for other commodities, all point to nominal and real wage stickiness
in the light of aggregate demand shocks as the principal explanation for the
existence of unemployment.
Keynes and Post Keynesians completely reject this neoclassical depiction
of an equilibrating labour market with self-adjusting characteristics. First,
Post Keynesians agree with other critics of orthodoxy for whom a labour
service cannot be conceived as just another commodity. As pointed out elsewhere (Seccareccia 1991), labour services have little in common with most
commodities, whether they be consumption or capital goods, to which the
standard tools of supply/demand are indiscriminately applied. Since it
cannot be stored, a labour service is a particular flow variable which, if left
unused, is lost for ever (Eichner 1979). Being instantaneously perishable,
the neoclassical price-auction model becomes highly inappropriate. In fact,
380
Wages and labour markets 381

the self-equilibrating role that inter-temporal arbitrage is supposed to play
in commodity markets is not only entirely absent in the market for labour
services but, because of the presence of hysteresis, may actually work
against the type of labour arbitrage analysed, for instance, by New Classical
theorists of the real business cycle. This is because, in addition to the complete loss of the current flow of labour services, unemployment also causes
the deterioration of skills, thereby making both workers and firms even
more anxious to contract during the current period.
Second, it has been recognized since Alfred Marshall that one cannot separate the labour service from the particular human being who oers it. In
opposition to the Walrasian price-auction model of the labour market that
presumes no prior relationship existing between buyers and sellers, the
money-wage bargain actually takes place in a social and historical setting in
which the relationship between employers and employees is an ongoing one.
Hence, the form of market coordination in line with the Post Keynesian perspective is that which accounts for the operation of customary practices and
social norms in regulating the level of money wages and their structure
across sectors. Compatible with the institutionalist labour market paradigm
(Gimble 1991), money contracts may be considerably insulated from market
forces and reflect primarily bargaining power and the normative pressures
arising from custom and workers beliefs as to what constitutes fairness in
the determination of wages.
Third, and perhaps most importantly, the whole notion of a distinct
aggregate labour market analysed in isolation must be abandoned in favour
of a more organic approach grounded on the principle of eective demand.
Unlike the market for other commodities, changes in the price of labour, the
average real wage, have both supply and aggregate demand eects which
impact on firms employment decisions. This is because wages are not
merely an element of a firms cost. Through their impact on a communitys
consumption, wages also directly influence aggregate demand. Hence, as
Keynes had made very clear, a reduction in money-wages will have no
lasting tendency to increase employment except by virtue of its repercussions either on the propensity to consume for the community as a whole, or
on the schedule of the marginal eciencies of capital, or on the rate of interest (Keynes 1936, p. 262). It is the demand-side eects of wage changes that
really matter to Post Keynesians, and these eects would normally work in
the opposite direction to what is generally theorized by neoclassical theory.
Moreover, a cut in wages may also further compound these contractionary
aggregate demand outcomes, which cannot be averted in a world of endogenous money via Pigovian wealth eects, because of the negative expectations of future market conditions (Deprez 1996). Consequently, following
Michal- Kalecki, Post Keynesians generally surmise a positively sloped
382
Wages and labour markets
aggregate demand curve for labour in the real wage/employment space

because of significant feedback eects via aggregate demand or, in some
cases, they suppose no relation whatsoever between real wages and employment (King 2001, p. 71).
Although rejecting the neoclassical notion of an aggregate labour
market, Post Keynesians point to the overwhelming significance of
demand, both the level and structure of aggregate demand, in conditioning
employment. While the aggregate level of demand sets the macroeconomic
constraints on the utilization of labour, it is the dual structure of product
markets in industrial economies traditionally typified by a core oligopolistic sector and a more peripheral competitive sector that defines the precise
segmentation of labour demand (Appelbaum 1979).
Following the work of institutionalist economists, the dominant oligopolistic sector is composed of megacorps (domestic or transnational) which,
because of their larger size, are commonly distinguished by more capitalintensive methods of production, and faced with a more stable demand
especially for skilled labour and a relatively better paid and more unionized
labour force. Indeed, technological requirements in the primary segment
call for more highly-skilled workers with firm-specific training, necessitating low labour turnover or employment variability. One crucial feature of
the primary sector is the existence of internal labour markets that regulate
internal mobility and promotion and are characterized by more rigid and
hierarchical wage structures patterned along formal seniority ladders. Such
internal labour markets are assumed to be largely insulated from the external labour market, except at the ports of entry via external job clusters and
wage contours. Internal labour markets meet a number of employer objectives, of which the most important is to guarantee that skills are transmitted vertically across seniority districts so as to minimize the cost of
on-the-job training while, at the same time, bringing about a more ecient
mutual supervision of workers once on the job. However, in response to
other competing corporate objectives, a wide diversity of internal labour
market systems can subsist that are more or less insulated from the external labour market (Grimshaw and Rubery 1998).
Although the primary sector can itself be subdivided into upper and
lower tiers, depending on such factors as the degree of capital intensity and
stable demand patterns, this sector must be formally distinguished from the
secondary sector. The latter, instead, normally comprises smaller, more
competitive firms, specialized in more labour-intensive activities and associated with lower wages and much greater employment flexibility. Indeed,
even where there are large corporations, as in some of the personal services
industries, work units tend to be small and isolated and thus, for instance,
create barriers to unionization. The relative absence of structured internal

labour markets in the secondary sector, however, is explained primarily by
firms need for lower-skilled employment, thereby entailing higher labour
turnover and greater sensitivity of wages to external demand pressures.
Given this particular fragmentation of demand, it follows that labour
supply for each segment is rooted in dierent social stratifications that are
perpetuated by the dynamic interaction between job characteristics and
social aliation a process termed by Eichner (1979) as human developmental or anthropogenic. Unlike neoclassical theory which presents the
labour supply decision within a static model of rational choice, Post
Keynesians focus on labour supply as an outcome of the cumulative acquisition of competencies, attitudes and habits that have been moulded by
workers social aliations and subcultures. Although habits and aliations
can be changed by formal education and on-the-job training, the labour
market tends generally to preserve distinct social strata of non-competing
groups.
While Post Keynesians recognize that wages may be aected by shortrun demand/supply considerations, wages are not the result of an optimizing process whose purpose is to allocate labour to its most productive
employment, as emphasized by neoclassical theory. As was well understood
by both classical writers and Marx, workers are the only input in production who must provide for their own sustenance and social regeneration.
Therefore, relative real wages have little to do with specific marginal products but reflect, instead, the social reproduction needs to maintain hierarchies within firms that have been sanctioned socially by force of custom;
this is dubbed by Rubery (1997) the social cohesion/social stratification
function of wages. Once a particular wage structure has evolved historically, customary norms lead to a process of calcification of these wage dierentials that generally can withstand market pressures and, hence, exhibit
a high level of empirical stability. Keynes, himself, had recognized the significance of this institutional stickiness of wage dierentials in the General
Theory when he attributed to workers the primary concern of protecting
their relative wages. Only during periods of acute crisis would such ossified
wage structures succumb to management pressures and result in new precedent-setting wage relativities.
In much the same way, the aggregate level of the money wage can essentially be considered a historical datum, serving as the calculable starting
point for actual bargaining between workers and employers. Given its ubiquity, the aggregate money wage presents itself as the anchor on which the
actual level of nominal values in the entire system depends. The aggregate
level of the real wage, on the other hand, is not such an arbitrary standard,
but represents primarily forces at work in the product market, especially
firms degree of monopoly in the primary sector. Unlike changes in the
384
Walrasian economics
money wage that would have their greatest influence on nominal magnitudes,
movements in the real wage would aect employment via their consequences
for aggregate eective demand. Through their feedback on the pattern of
spending, real-wage movements play a key role in the determination of
overall employment, as well as its allocation between the primary and secondary sectors of the labour market. Accordingly, high real-wage growth
would engender a virtuous cycle of prosperity as well as an increase in the
share of the primary sector of the labour market a scenario characteristic
of much of the early post-1945 golden age of Western capitalism. Conversely, a fall in the real wage, associated with a concomitant rise in the
Kaleckian degree of monopoly, would generate a vicious cycle characterized
by a growing problem of eective demand and a proliferation of low-wage
jobs in the secondary sector of the labour market.
M S
See also:
Eective Demand; Employment; Institutionalism; Marginalism; New Keynesian Economics;
Unemployment.
References
Appelbaum, E. (1979), The labor market, in A.S. Eichner (ed.), A Guide to Post-Keynesian
Economics, White Plains, NY: M.E. Sharpe, pp. 100119.
Deprez, J. (1996), Davidson on the labour market in a monetary production economy, in P.
Arestis (ed.), Keynes, Money and the Open Economy: Essays in Honour of Paul Davidson,
Vol. 1, Cheltenham, UK: Edward Elgar, pp. 12343.
Eichner, A.S. (1979), An anthropogenic approach to labor economics, Eastern Economic
Journal, 5 (3), 34966.
Gimble, D.E. (1991), Institutionalist labor market theory and the Veblenian dichotomy,
Journal of Economic Issues, 25 (3), 62548.
Grimshaw, D. and J. Rubery (1998), Integrating the internal and external labour markets,
Macmillan.
King, J.E. (2001), Labor and unemployment, in R.P.F. Holt and S. Pressman (eds), A New
Guide to Post Keynesian Economics, London and New York: Routledge, pp. 6578.
Rubery, J. (1997), Wages and the labour market, British Journal of Industrial Relations, 35
(3), 33766.
Seccareccia, M. (1991), An alternative to labour-market orthodoxy: the Post-Keynesian/institutionalist policy view, Review of Political Economy, 3 (1), 4361.
Walrasian Economics
Walrasian economics originated in the work of Lon Walras (1874 [1926])
and was one of several neoclassical approaches emerging from the marginal
revolution in the late nineteenth century. Its hallmark was the focus on the
interrelations of markets and their simultaneous, or general, equilibrium.
Walrasian economics 385

Walrasianism was, therefore, distinct from Marshallian neoclassicism, which
sought analytic devices to neutralize interdependencies so as to allow partial
equilibrium analysis of particular sectors of the economy. And it was dierent, too, from the neoclassicism of J.B. Clark, who resorted to extensive
aggregation to avoid the complexities resulting from interdependence. Since
Walrass early formulation, his general equilibrium theory has been increasingly refined, and it became the dominant form of orthodoxy during the
twentieth century, only recently being somewhat displaced by game theory.
The neoclassical quality of Walrasian economics is evident in the derivation of agents choices, or supplies and demands, from particular types of
maximization. It is assumed that each consumers domain of choice (commodity space), preferences (typically represented as a utility function) and
assets (labour capacities, physical possessions and financial securities), and
each producers technology (inputoutput combinations) are exogenously
specified, and economic interaction takes place only through competitive
markets. Every consumer is depicted as maximizing utility subject to a
budget constraint, and every producer maximizes profit constrained only
by technology. A set of prices that allows these optimizations to be realized
simultaneously, so making demands and supplies compatible, and markets
clear, is a Walrasian equilibrium.
Two varieties of Walrasian economics can be distinguished: Arrow
Debreu intertemporal equilibrium theory and temporary equilibrium analysis. They dier in their treatment of time: that is, in how agents relate to
the future. The ArrowDebreu version classifies commodities not only by
their physical properties and locational attributes, but also by their date of
availability and the state of the world, which specifies the corresponding
values of variables that were previously uncertain. The rationale for this is
straightforward. All of these characteristics can aect utilities and profits,
and therefore demands and supplies. For example, consumers preferences
are generally sensitive to a commoditys physical properties and the location
where consumption occurs, as well as embodying time preferences and attitudes towards risk. But the implications of conceptualizing commodities in
this fourfold way are dramatic. When coupled to the usual assumption that
agents can completely rank consumption bundles and inputoutput combinations, it means that there will be a single decision date for all agents and
a comprehensive set of futures markets and contingent commodity, or insurance, markets. Agents will form supplies and demands for every date and
for every contingency, and will determine all of their choices in the very first
period.
Walrasian temporary equilibrium theory implicitly recognizes the unreasonableness of assuming that agents have complete rankings of commodity bundles. Instead, they are likely to be uncertain as to what commodities
386
Walrasian economics
and technologies will be available in the future and what their preferences
will be, as well as the exact composition and value of their assets. And typically they will encounter transaction costs that are especially high for
trading in futures markets and contingent commodities. Thus markets will
be incomplete, and at least some trading will take place sequentially. An
ArrowDebreu equilibrium will be impossible, and any equilibrium will be
temporary. J.R. Hicks (1946) formalizes the matter as follows. There are
a number of periods and, at the beginning of each, spot markets exist
together with some futures markets and contingent commodity markets.
Agents can trade on the markets that exist, and will do so on the basis of
expectations about future prices for commodities that are not presently
tradable. Equilibrium is defined only in terms of clearance on those
markets that exist, and these are not comprehensive. At the beginning of
the next period, markets reopen, and new trading can occur on spot
markets and newly available markets for forward contracts. Expectations
formed at the beginning of the first period may turn out to be incorrect, and
will be revised, so that market prices in the second period will reflect new
expectations. At the start of the third period markets reopen again. And so
on. The economy thus moves in a sequence of temporary equilibria.
The causation structure of both types of Walrasian theory is clear-cut.
The equilibrium values of the endogenous variables (prices and quantities
traded) are determined by the exogenously specified preferences, asset
endowments, technologies, optimizing behaviours and agents expectation
formation rules. The endogenous variables, prices and quantities are, of
course, the microeconomic components of the macroeconomic variables.
The time path of these aggregates can be very complex in both forms of
Walrasian model. They certainly need not result in the steady states of
Robert Solows growth model. The causal fundamentals would have to take
very special forms for this to occur. However, no matter how complicated,
equilibria will always involve market clearance, so there will be no involuntarily unemployed labour or underutilized productive capacity in the
Keynesian sense of these terms. (There can be excess supplies in Walrasian
equilibria, providing there is free disposal, but they will correspond to zero
prices.)
Five problems have been examined in both types of Walrasian theory.
First, under what conditions do equilibria exist? Second, when will the
economy be stable, in the sense of disequilibrium proving transitory? Third,
what are the eciency properties of equilibria? Fourth, what circumstances
will guarantee a unique equilibrium, and, fifth, generate definite comparative static results? Overall, coordination through prices as depicted by
Walrasianism is shown to be a delicate matter. Very restrictive assumptions
are required to ensure existence, stability, eciency, uniqueness and definite
Walrasian economics 387

comparative statics. For example, take the existence question, that is, the
problem of determining what types of consumers preferences, producers
technologies, asset distributions and expectation formation mechanisms
will ensure that there is a vector of prices that will clear all markets simultaneously. Since Walrasianism is essentially a set of propositions about the
properties of equilibria, the significance of the problem is dicult to exaggerate. But the assumptions required to guarantee existence are stringent.
In particular, they must ensure that demands and supplies vary continuously with prices. Further restrictive assumptions are required to guarantee
that, if equilibrium exists, it is unique, and stable, and that parameter shifts
engender definite changes in endogenous variables. This has led some
eminent orthodox theorists to suggest that Walrasianism is not the route to
proceed along in analysing real economies, or, alternatively, that it should
quickly lead into a more secure path (Arrow and Hahn 1971).
Such caution is notably absent from orthodox macroeconomics, whether
monetarist, New Classical, or real business cycle theory. Here actual economies are modelled as oscillating between two types of Walrasian temporary equilibrium, those with rational expectations and those deviating from
rational expectations. The deviations may be the result of monetary or real
shocks, and are regarded as quickly self-correcting. Walrasian equilibria
thus prevail continuously and are each treated as unique, stable and exhibiting definite comparative statics. Since it is dicult to justify any of this in
terms of the theorems proved by Walrasian theorists, orthodox macro
economists usually take refuge within the confines of Milton Friedmans
instrumentalist methodology to justify their extreme modelling simplifications, most notably their resort to heroic aggregation, where markets are
reduced to a very small number and where there is a single economic agent
with well-behaved preferences and constraints. In such a context, the problems of existence, stability and eciency are much reduced, and unambiguous comparative static results are easier to generate.
Beginning in the 1960s, some Keynesians insightfully exposed the key
assumption guaranteeing market clearance as a property of equilibria in
Walrasian theory. Clower (1965) and Leijonhuvud (1968) argued that this
resulted from treating agents as formulating their demands and supplies in
the belief that they can always trade whatever quantities they desire, providing only that they deliver equivalents in exchange. This seems innocuous, but is in fact crucial, as can be appreciated by considering a Walrasian
disequilibrium. Obviously, in such a situation not all agents choices can be
realized simultaneously, and some will be rationed if trades actually occur.
Then it is not unreasonable to imagine that this rationing will aect choices.
For example, a consumer who is quantity constrained in the sale of labour
and expects this to continue at future dates will not necessarily alter the
388
Walrasian economics
supply of labour from that resulting from a Walrasian maximization of

utility, but will probably reduce consumption demands, thereby contributing to an eective demand deficiency. A producer who is quantity rationed
in the sale of output, and expects this to continue at future dates, will not
necessarily reduce the supply of output below the Walrasian level, but is
likely to reduce demand for labour inputs, again contributing to a shortfall
in demand. Agents supplies and demands will not, therefore, be correctly
specified by Walrasian theory, and their equilibration may involve significant deviations from market clearance. If a new equilibrium occurs it can
be Keynesian, in the sense of there being excess supplies of commodities,
including the involuntary unemployment of labour, without the corresponding prices being zero. Furthermore, disequilibria will exhibit multiplier processes, which are wholly absent from Walrasian economics.
Post Keynesians have been very much more critical of the entire
Walrasian edifice, including all notions of optimization and equilibrium.
Radical uncertainty, it is claimed, is not only something irreducible to a
probabilistic calculus; it cannot be sensibly treated as aecting only the way
expectations are formed. Instead, it will result in behaviour that cannot be
modelled as maximization subject to constraints. And the institutional
structure in which decisions occur will also be significantly aected, particularly in the monetary and financial sectors. The historical processes consequent upon both are unlikely to be representable as equilibria because
speaking in Walrasian terms the determinants of equilibria will undergo
change in the process itself. Thus the notion of general equilibrium is irrelevant to a world of ignorance and uncertainty, where irreversible decisions
must be taken in calendar time and equilibrium states are (if indeed they
are attained) invariably path dependent (King 1995, p. 245).
M.C. H
See also:
Bastard Keynesianism; Equilibrium and Non-Equilibrium; Marginalism; Says Law.
References
Arrow, K.J. and F.H. Hahn (1971), General Competitive Analysis, Edinburgh: Oliver & Boyd.
Clower, R.W. (1965), The Keynesian counter-revolution: a theoretical appraisal, in F.H.
Hahn and F. Brechling (eds), The Theory of Interest Rates, London: Macmillan, pp. 10325.
Hicks, J.R. (1946), Value and Capital, Oxford: Oxford University Press.
King, J.E. (1995), Conversations with Post Keynesians, London: Macmillan.
Leijonhuvud, A. (1968), On Keynesian Economics and the Economics of Keynes, Oxford:
Oxford University Press.
Walras, L. (1874), Elements of Pure Economics, Edition Dfinitive, 1926, London: Allen &
Unwin, 1954.
Name index
Akerlof, G. 113
Anderson, W.L. 229
Arestis, P. 14, 85, 96, 116, 158, 217,
34650
Arrow, K.J. 3856
Asimakopulos, A. 3067, 328
Atkinson, A.B. 46, 328
Azariadis, C. 113
Coddington, A. 15960
Colander, D. 188
Cornwall, J. 945, 3227
Cornwall, W. 945, 27580
Courvisanos, J. 1916
Crafts, N. 19
Cross, R. 94
Crotty, J.R. 254
Bagehot, W. 57
Baily, M.N. 113
Baran, P.A. 195, 371
Barro, R.J. 271
Bauer, O. 371
Bausor, R. 130
Bell, S. 2428
Bellamy, E. 317
Belloore, R. 60
Beveridge, W.H. 155, 157, 269
Bhaduri, A. 173, 373
Bhaskar, R. 823
Blair, T. 338
Blanchard, O. 113, 114, 233, 278
Blatt, J. 178
Blecker, R. 200205
Bloch, H. 24952
Bhm-Bawerk, E. von 5
Bowles, S. 182
Brid, J.C.M. 175
Brown, A. 826
Brumberg, R. 73
Bunting, D. 727
Burbidge, J. 328
Davidson, P.
on axiom of gross substitution 70,
233
on Austrian economics 68
on balance-of-payments-constrained
growth 19
on central banks 59
on critical realism 85
on eective demand 22937, 253,
255
on employment 255
on ergodicity 97, 15960, 2367,
28083, 36870
on nance 28, 89, 208, 247
on international nance 33, 1345,
204, 260, 346
on investment 2078
and Journal of Post Keynesian
Economics 21518
on Keyness General Theory 22937,
269
on methodology 85, 96
on money 93, 94, 198, 2334, 236,
247, 369
on the multiplier 2689
on New Keynesian economics 230,
233
on rational expectations 280
on Says Law 23035
on the Tobin tax 346
on uncertainty 26670
Day, R. 95
Deane, P. 50
Debreu, G. 3856
Delli Gatti, D. 945
Dixon, R.J. 174
Carabelli, A. 160, 364

Carvalho, F. 5760
Cencini, A. 60
Chamberlin, E.H. 222
Champernowne, D. 46
Chandler, A. 167
Chiarella, C. 95, 179
Chick, V. 28, 59, 268, 290
Clark, J.B. 296, 385
Clinton, W.J. 338
Clower, R.W. 387
389
390
Name index
Dobb, M.H. 47, 50, 212

Domar, E.S. 135, 269, 3723
Douglas, C.H. 371
Dow, A. 247
Dow, S.C. 1115, 223, 78, 81, 967,
247
Downward, P. 96101, 252
Duesenberry, J. 2078
Dunlop, J.T. 374
Dunn, S.P. 28084
Dutt, A.K. 189, 28993
Dymski, G. 256
Eatwell, J. 322
Edgeworth, F.Y. 162
Eichner, A.S. 68, 96, 167, 218,
383
Eisner, R. 153
Engels, F. 314
Fazzari, S. 208, 2527
Feynman, R.P. 1112
Fine, B. 517
Fischer, S. 276
Fisher, I. 146, 187, 206, 229
Fitzgibbons, A. 160, 2715
Flaschel, P. 95, 179
Foley, D.K. 94
Fontana, G. 60, 23741
Forstater, M. 3749
Franke, R. 95
Freeman, C. 192
Friedman, M. 27, 73, 156, 187, 221,
259, 2712, 387
Frisch, R. 39
Fusfeld, D. 217
Galbraith, J.K. 21617
Gallegatti, M. 95
Garegnani, P. 93
Garretson, H. 256
Georgescu-Roegen, N. 70
Gerrard, B. 15964
Giddens, A. 338
Gintis, H. 182
Glickman, M. 36670
Gnos, C. 60
Godley, W. 38
Gomulka, S. 193, 330
Goodhart, C.A.E. 262
Goodwin, R.M. 4041, 49, 95, 176,

178, 185
Graziani, A. 60, 1425
Grossman, H. 226
Haavelmo, T. 98
Hahn, F.H. 234
Halevi, J. 92
Hall, R.L. 251
Hansen, A.H. 25
Harcourt, G.C. 29, 4451, 192, 216
Harvey, J.T. 1315
Harrod, R.F. 17, 19, 25, 40, 47, 49, 92,
135, 17074, 176, 212, 250, 3723
Hawtrey, R.G. 339
Hayek, F. von 59, 92, 2212
Heilbroner, R. 217
Hendry, D. 98
Henry, J.F. 3416
Hewitson, G. 2024
Hicks, J.R. 17, 25, 222, 223, 276, 283,
284, 342, 3558, 386
Hicks, U. 221
Hitch, C.J. 251
Hobson, J.A. 3712
Hodgson, G.M. 45
Howard, M.C. 3848
Howells, P. 25761
Hudson, J. 11217
Hume, D. 202
Hymer, S. 167
Jevons, W.S. 216, 249
Jorgenson, D. 252, 255
Kahn, R.F. 47, 489, 211, 216, 221,
230, 266, 304
Kahnemann, D. 132
Kaldor, N.
on business cycles 40
on capital theory 49
on central banks 589
on cumulative causation 94, 174
economics of 2216
on equilibrium 222
on expectations 135, 138
on growth theory 49, 92, 1256,
1712, 1767, 2235
on income distribution 1712, 177,
1834, 2234
Name index 391

on innovation 193
on money 589, 264
on the multiplier 3067
on speculation 138
on taxation 2223
Kalecki, M.
on budget decits 347
on business cycles 39, 41, 93, 193,
329
on economic growth 923, 102,
1934, 330, 3578, 372
on economic policy 1012
economics of 2269
on government spending 357
on income distribution 49, 173, 183,
2267, 329, 3734
on innovation 1914, 209
on investment 2059
and Joan Robinson 212
microeconomics of 106, 109, 183,
251, 29091
on monopoly capital 2278
and paradox of costs 173
on political business cycles 41, 151,
157, 377
and the principle of increasing risk
146, 207
on prots 148, 2267, 2967
on the rate of interest 2289
on socialism 227, 358
on the traverse 3568
on uncertainty 228
on underconsumption 372
on wages 3734, 381
Katzner, D.W. 12631, 217
Keen, S. 95, 17580
Kenyon, P. 192
King, J.E. 37074
Kirzner, I. 5, 7
Kiyotaki, N. 113
Klein, L. 227, 281
Knapp, G.F. 263
Knight, F.H. 283, 296
Kornai, J. 178
Kotliko, L. 328
Kregel, J.A. 59, 94, 135, 139, 177, 247
Kriesler, P. 92, 251, 3559
Krugman, P. 19
Kurz, H.D. 93
Kuznets, S. 230
Lachmann, L. 5, 7
Lancaster, K. 69
Lange, O. 25
Laramie, A.J. 32832
Lavoie, M. 22, 60, 6972, 80, 92, 193,
2467, 374
Lawson, T. 13, 825, 96, 160, 282
Leamer, E. 99
Lee, F.S. 218, 251, 2859
Le Guin, U. 317
Leijonhufvud, A. 27, 230, 387
Lekachman, R. 153
Lerner, A.P. 102, 149, 151, 188, 216,
222, 251
Lodewijks, J. 2430
Lpez, G.J. 348
Lotka, A. 178
Lowe, A. 3569, 377
Lucas, R.E. 27, 253, 271
Lutz, M.A. 69
Lux, K. 69
Luxemburg, R. 191, 226, 372
Machlup, F. 5, 249, 251
Mair, D. 32832
Malthus, T.R. 44, 154, 304, 309, 311,
370
Mankiw, N.G. 29, 113
Marcuzzo, M.C. 211215
Marglin, S. 373
Marshall, A. 445, 50, 94, 106, 110,
211, 214, 216, 249, 290, 295, 341,
381
Marx, K.
on business cycles 41, 154, 178
and the Cambridge economic
tradition 49
and classical political economy 319
and growth theory 116, 178, 213,
269
inuence on Joan Robinson 212,
214, 372
and the monetary theory of
production 264, 311, 344
and the rate of interest 302
and Says Law 309, 31112
on technical progress 170, 213
on unemployment 41, 116, 154,
377
392
Name index
Marx, K. (continued)
and value theory 56
on wages 383
Matthews, R.C.O. 50
Matzner, E. 33741
McCombie, J.S.L. 1520, 175
McGregor, P. 18
McKenna, E.J. 15
McKinnon, R. 87, 89
Meade, J.E. 25, 46, 478, 306
Mearman, A. 98100
Medio, A. 95
Menger, C. 56, 249
Messori, M. 60
Milberg, W. 16570
Milgate, M. 322
Miliband, R. 314
Mill, J. 309
Mill, J.S. 216, 309
Minsky, H.P.
on banking 224, 148
on budget decits 37, 147
on business cycles 40, 81, 94, 1467,
219
on central banks 59
on disequilibrium 28
on economic growth 179
on economic policy 37, 1478, 378
on employer of last resort policy 37,
378
on expectations 135, 1368
on nancial fragility 24, 78, 88,
1459, 219
on the nancial instability
hypothesis 1459
on investment 1368, 2078, 2556
on money 345
Mises, L. von 5, 6
Mitchell, W.F. 1539
Modigliani, F. 25, 73, 173, 252, 255
Mongiovi, G. 31222
Moore, B.J. 21, 59, 11721, 2467,
264, 267
Morgenstern, O. 5
Moszkowska, N. 371
Mott, T. 92, 20510
Mundell, R. 262
Musgrave, R.A. 328
Muth, J. 273
Myrdal, G. 217, 222
Nell, E.J. 68, 93, 176, 178

Nevile, J.F. 14953
Newton, I. 11
ODonnell, R.M. 160, 35965
OHara, P.A. 21520
Ohlin, B. 242
Okun, A. 276, 333
Palley, T. 1816
Panico, C. 17075, 321
Pareto, V. 249
Parguez, A. 60
Parkin, M. 113
Parsons, S.D. 510, 85
Pasinetti, L.L. 4950, 68, 93, 11112,
171, 1767, 1834, 377
Patinkin, D. 230, 269
Perez, C. 192
Peterson, W. 217
Petty, W. 153
Phelps, E.S. 113, 156, 276
Pigou, A.C. 46, 211, 222
Pivetti, M. 299303, 321
Pollin, R. 3049
Popper, K. 213
Poulon, F. 60
Pressman, S. 196200
Puu, T. 95
Ramsey, F. 162, 364
Realfonzo, R. 6065
Reddaway, W.B. 25
Ricardo, D. 47, 50, 55, 202, 21213,
214, 217, 294, 297, 300, 304,
30911, 319
Rider, C. 35054
Robbins, L. 221
Robertson, D.H. 46, 60, 142, 242, 249
Robinson, E.A.G. 47
Robinson, J.
on Bastard Keynesianism 25
on business cycles 40
on capital theory 51, 95, 1812, 213,
250
on comparative advantage 167
on economic growth 4950, 1767,
21213, 250, 355, 372
economics of 21115
on equilibrium 214
Name index 393

on expectations 135
on imperfect competition 47
on ination 1878
inuence of Kalecki on 49, 212,
227, 372
on methodology 21314, 268
on the new mercantilism 2012
on the rate of interest 300
on time in economic theory 92, 95,
214, 355
Rochon, L.-P. 60, 1459
Roemer, J.E. 182
Romer, D. 92
Rosser, J. Barkley, Jr. 925, 218
Rotenberg, J. 113
Rowthorn, R.E. 189
Sadigh, E. 60
Salter, W.E.G. 192
Salvadori, N. 93
Samuels, W.J. 217
Samuelson, P.A. 257, 29, 40, 171,
261, 280
Sardoni, C. 30913
Sargent, T. 271
Savage, L. 282, 283
Sawyer, M.C. 92, 1015, 158, 218, 346
Say, J.-B. 154, 30911
Schmitt, B. 60, 267
Schmoller, G. 5
Schmookler, J. 193
Schumpeter, J.A. 5, 60, 62, 95, 114,
191, 192, 238, 297
Seccareccia, M. 38084
Seidman, L.S. 3328
Semmler, W. 95
Settereld, M. 94, 10512
Shackle, G.L.S. 5, 28, 129, 135,
15960, 217, 281, 283, 298, 3667,
369
Shapiro, C. 113
Shapiro, N. 657, 218
Shaw, E.S. 47
Sherman, H.J. 31318
Shove, G. 47
Silverberg, G. 95
Sismondi, J.C.L. Simonde de 154, 309
Skidelsky, R. 199
Skott, P. 3843, 80, 94
Slutsky, E. 39
Smith, A. 94, 170, 202, 213, 214,
30910, 319
Smithin, J. 94, 18691
Solow, R.M. 27, 29, 92, 113, 171, 213,
281, 296, 386
Sraa, P.
on capital theory 51, 556, 171, 213,
318
and classical political economy 214,
302, 319, 356
economics of 31822
and eective demand 31920
and growth theory 178
on income distribution 182, 31819,
3212
and Joan Robinson 211, 21314
on money 3212
and Post Keynesian economics 93,
31921
on the rate of interest 173, 3023,
321
on the rate of prot 173, 182
on the theory of production 295,
356
Stalin, J.V. 317
Steedman, I. 172
Steindl, J. 37, 92, 173, 1956, 209, 228,
372, 377
Sterman, J. 95
Stiglitz, J.E. 77, 328
Stohs, M. 11
Stone, R. 44
Studart, R. 8791
Summers, L. 114, 328
Swales, K. 18
Swan, T. 92
Sweezy, P.M. 195, 228, 371
Taylor, J.B. 121, 187, 259, 276
Thirlwall, A.P. 16, 1745, 217,
2216
Thornton, W. 242
Tinbergen, J. 27
Tobin, J. 29, 333, 346, 348
Toporowski, J. 2269
Trigg, A.B. 26570
Tugan-Baranovsky, M. 226, 372
Tversky, A. 132
Tymoigne, E. 13541
394
Name index
Urban, F.M. 162

Van Ees, H. 256
Varga, E. 371
Variato, A.M. 256
Veblen, T. 116, 197
Vernengo, M. 3034
Vernon, R. 204
Vickers, D. 217
Volterra, V. 178
Wallace, N. 271
Wallich, H. 188, 333
Walras, L. 249, 384
Walters, B. 85
Watts, M.J. 1539
Webster, E. 2948
Weintraub, S. 25, 27, 188, 21516, 218,

3337, 373
Weiss, A. 77
White, H.D. 31
Wicksell, K. 60, 187, 301
Wieser, F. von 5
Wimsatt, W.C. 11
Winnett, A. 1216
Wolfson, M.H. 7782
Worswick, G.D.N. 227
Wray, L.R. 37, 102, 247, 2615
Yellen, J. 113
Young, A.A. 94, 221
Young, D. 85
Zannoni, D. 15
Subject index
Absolute advantage 203
Accelerator principle 207, 372
Accumulation of capital see growth
Accumulation of Capital, by Joan
Robinson 21213
Administered pricing see mark-up
pricing; oligopoly
A General Theory of the Price Level, by
Sidney Weintraub 333
Agency 15, 97, 125, 1323, 1967,
252, 2545, 2812, 284, 364,
3845
Aggregate supply and demand model
278, 10512, 13541, 23035,
291, 373
Animal spirits 39, 42, 141, 163, 198,
209, 235, 254, 274, 344
Asset prices 147, 340
Asymmetric information 23, 77, 79,
87, 114, 230, 256, 279, 347
Atomism 13, 967, 162
Austrian economics 510, 84, 221,
283, 297, 356
Babylonian methodology 1115, 85,
967
Balance-of-payments-constrained
growth 1520, 94, 104, 15052,
158, 1745, 2045, 225
Banana parable, of J.M. Keynes 240
Bandwagon eects 133
Banking
circuit theorists on 614, 2678
credit rationing by 7882
and development nance 88, 90
and endogenous money 11821
and interest rates 11921, 247
Keynes on 58, 1424, 240, 307
main entry 2024
Minsky on 146
and the theory of money 2634
see also central banking; money
Barter 262, 31011
see also neutrality of money
Bastard Keynesianism 2430, 292
see also neoclassical economics;

neoclassical synthesis
Beveridge Report 222
Borrowers risk 78, 1389, 207
Bretton Woods 3034, 48, 102, 134,
166, 202, 204
Budget decits 289, 348, 102, 147,
14953, 158, 273, 331, 348
Business cycles
banks and 23
dynamics of 925
globalization and 166, 169
growth and 95, 176, 178
Kaldor on 222
Kalecki on 2267, 229
Keynes on 23, 23841
main entry 3843
Minsky on 1458, 219, 229
Cambridge circus 211
Cambridge controversies see capital
theory
Cambridge economic tradition 4451,
517, 845
Cambridge equation 1712
Cambridge Journal of Economics 215
Capacity utilization 1734
Capital accumulation see growth
Capital markets 8791
Capital movements, international
312, 42, 1315, 1656, 2014,
3469
Capital theory
Cambridge critique of 49, 93, 181,
209
and environmental economics
1245
and income distribution 1812
Joan Robinson on 181, 213, 250
main entry 517
neoclassical 171, 2812, 2967
Sraa on 171, 293, 31822
Capitalisms Ination and
Unemployment Crisis, by Sidney
Weintraub 3346
395
396
Subject index
Central banks
and endogenous money 118, 120,
264
and interest rates 21, 108, 120, 257,
264
and international nance 103
Keynes on 307
as lender of last resort 24
main entry 5760
Marshall on 45
Minsky on 147
see also monetary policy; Taylor rule
Chaos theory 42, 282
Choice of technique 516
Circuit theory 615, 119, 264, 267
Class conict 171, 1845, 3734
Classical dichotomy 153, 258
see also neutrality of money; Says
Law
Classical economics, J.M. Keynes on
2327
Classical political economy 44, 72,
17074, 212, 271, 289, 300, 302,
30911, 319, 3567 see also
Marxian economics; post-classical
economics; Sraan economics
Classical Keynesianism, by Sidney
Weintraub 336
Closed-system thinking 1213, 834,
96, 282
Cobweb theorem 225
Co-integration 281
Cold War 30
Commodity money 117, 261, 263
Communism 35054
Comparative advantage 203
Competition 657, 1679
Complex dynamics 945, 177, 179
Conict theory of ination 18990,
219
Consumer theory 6872, 73, 197
Consumption 727, 197, 37074
Conventions 23, 13, 23, 68, 80, 133,
14041, 163, 254, 312
Costs 2858, 2945, 300302
Cost-plus pricing see mark-ups;
oligopoly; pricing and prices
Cost-push ination 33, 93, 18890
see also tax-based incomes policy;
wages policy
Credit money 212, 11721, 264

Credit rationing 20, 223, 63, 7882,
91, 120, 279
Crises see business cycles
Critical realism 13, 826, 967, 100
Crowding-out 36, 14950
Cumulative causation 174, 217, 359
Debt 15, 224, 64, 1468, 1528,
344
Debt deation theory of crises 147,
229
Decits see budget decits
Degree of monopoly 173, 183, 251,
329, 3834
Demand, consumer 6872
Demand-pull ination 18990
Demi-regs 834
Democracy 31317
Depreciation 3045, 329
Deregulation 3378, 376
Development economics 227, 315
Development nance 8791
Dictatorship 31317
Disequilibrium 28, 123, 1512, 3559,
3878
see also equilibrium; history and
equilibrium
Distribution see income distribution
Dual labour markets 3824
Dualism, philosophical 1314
Dynamics 925, 12830, 171, 17580,
212
see also business cycles; growth;
traverse
Econometrics 1718, 27, 96101,
11415, 281, 364
Economic growth see growth
Economic Philosophy, by Joan
Robinson 21314
Economic policy
Austrians on 89
and economic development 91
and environmental issues 122
and employment 1558, 375, 3789,
380
institutionalists on 179200
international aspects 1345, 167,
2015
Subject index
and 219
Kaldor on 172
Kalecki on 348
Keynes on 45, 48, 24041
main entry 1015
Minsky on 1467
New Classical economists on 2715
New Keynesians on 28, 276
and saving 308
and stagation 3257
and transition economies 354
uncertainty and 369
and unemployment 375, 3789
see also scal policy; monetary
policy; taxation; tax-based
incomes policy; Third Way; Tobin
tax; wages policy
Economics, by Paul Samuelson 26
Economics of Imperfect Competition,
by Joan Robinson 47, 211, 222
Eective demand
and cumulative causation 217
and growth 178
Kalecki on 358
Keynes on 31, 136, 1389, 162, 231,
2367, 250
main entry 105112
and microfoundations 255
and production 2923
Sraa and 31920
and the Third Way 340
in transition economies 353
and unemployment 116, 269, 2923,
3757, 381
see also Keyness General Theory;
Says Law
Eciency wage theory 11314, 2767,
292
Emergent properties 4
Employer of last resort, state as 37,
102, 1578, 378
Employment 10510, 11217, 1369,
1539, 3734, 3748, 38084
see also labour market;
unemployment; wages
Endogenous growth theory 56, 92, 94,
125, 224
Endogenous money
banks and 20
397
Dow on 789
and scal policy 150
Kalecki on 228
Keynes on 28, 2467
main entry 11721
and monetary policy 257
and monetary theory 2645
and the rate of interest 108, 302,
321
see also circuit theory; horizontalism
Entrepreneur economy 88
Entrepreneurs 78, 62, 13441, 144,
23940, 2967
Environmental economics 72,
11721
Equality 313, 31617
Davidson on 28, 369
and scal policy 1512
Harrod on 177
Hicks on 356, 386
Joan Robinson on 159, 21214,
355
Kaldor on 94, 177, 2223
Kalecki on 3579
Keynes on 136, 1634, 250
in labour market 250
Lowe on 3568
main entry 12631
Marshall on 45
Minsky on 28
in Sraan economics 359
and time 21314, 3423
temporary 3868
and Tobin tax 347
in Walrasian economics 3858
Ergodicity 97, 140, 2335, 28084,
3689
Essay on Marxian Economics, by Joan
Robinson 212
Essays in the Theory of Employment,
by Joan Robinson 213
Essays in the Theory of Economic
Growth, by Joan Robinson 212
Ethics and economics 3645
Excess capacity 178
see also capacity utilization
Exchange rates 1517, 3034, 1023,
120, 1315, 151, 158, 166, 2014,
260, 349
398
Subject index
Expectations
asymmetric 79
Austrian economists on 67,
103
and credit rationing 789, 81
Davidson on 28081
and equilibrium 127
and exchange rates 1323
and nancial fragility 78
and investment 2056
Kalecki on 228
Keynes on 161, 163, 2446, 290
and liquidity preference 2446
and Lucas critique 253
main entry 13441
in neoclassical economics 2534,
342
rational 2534, 2734, 280, 284
and stagation 3867
see also New Classical economics;
uncertainty
Exploitation 182
Export-led growth 174, 202
Export multiplier 1719
Externalities 123
Fiat money 117, 263
Finance 132, 1429, 208, 2556, 267,
295, 3068, 3469
Finance motive for holding money 62,
1425, 247
Financial crises 81, 216, 219
Financial fragility 78, 91, 146
Financial innovation 187
Financial instability hypothesis 4041,
8891, 1459
Fiscal policy
and Bastard Keynesians 26, 28
Chick on 28
and economic policy 1012
Godley on 38
and ination 32
institutional economists on 199200
Kaldor on 174
Kalecki on 34, 357, 32832
main entry 14953
Minsky on 37, 147
and the multiplier 2656
in New Classical economics 2723
and socialism 317
Steindl on 378
see also budget decits; taxation
Forced saving 239
Foreign direct investment 1678
Formalism in economics 412, 46,
364
see also econometrics
France 60
Free trade 167, 201, 214
Full-cost principle 2512
see also mark-ups; pricing and prices
Full employment
Davidson on 369
and employer of last resort 378
and scal policy 1015, 149, 151
globalization and 1667
Hicks on 356
and ination 102
Joan Robinson on 212
Kaldor on 177
Kalecki on 367, 101, 151, 377
Keynes on 101, 166, 246, 3767
Lowe on 3567, 377
main entry 1539
in neoclassical economics 26, 242,
250, 380
and the traverse 3567
see also eective demand; Says Law;
unemployment
Functional nance 102, 149
Fundamentalist Keynesians 15964
Funding 889
Game theory 385
General equilibrium theory 123, 261,
342, 3848
General Theory of Employment,
Interest and Money, by J.M.
Keynes
borrowers and lenders risk in 78,
1389
on expectations 135, 138, 163
on full employment 169, 250
on income distribution 169
on investment 2056, 21718, 290,
306, 319
on the labour market 250
Subject index
on liquidity preference 2467
main entry 22937
and the monetary theory of
production 344
on money 1424, 2467
and non-ergodicity 283
on the rate of interest 2467, 301,
307
on saving 3067
on Says Law 311
on speculation 346
Sraa on 31920
and Treatise on Money 238, 241
on uncertainty 15960, 1623,
21718
and wages 31920
German historical school of
economics 283
Gibson paradox 300
Globalization 16570, 3389
Golden age 30, 323, 49, 1767, 213,
3236
Government expenditure 348,
14953, 2656, 317
see also scal policy
Government policy see economic
policy
Great Depression 147, 154, 170,
1789, 242, 371
17075, 177, 2234
Growth, economic
balance of payments and 1520,
94, 104, 1745
endogenous 15
Harrod on 17, 47, 49, 269, 3723
Kaldor on 49, 1712, 174, 2235
Kalecki on 93, 104, 3589, 372
main entry 17580
neoclassical theory of 16, 545,
1256
neo-Ricardian theory of 174
Pasinetti on 93, 1712
Steindl on 209
see also dynamics
History versus equilibrium 21213,
281, 3415, 3559
399
Holism 2
Horizontalism 213, 59, 789, 121,
2645
Human rights 379
Hysteresis 11315, 137, 359, 381
Imperfect competition 21112, 222,
255, 277
see also degree of monopoly;
monopoly; oligopoly
Income distribution
in classical political economy 300,
3023, 31819
and scal policy 152
and growth 17075
Kaldor on 1712, 2234
Kalecki on 2023
Keynes on 239
main entry 1816
monetary policy and 259
neoclassical theory of 517
and socialism 31316
Sraa on 31819
and underconsumption 37074
Implicit contracts 113
Income and substitution eects 712
Incomes policy see wages policy
Inequality 338, 340
Ination
Austrian economists on 7
and the balance of payments 151
and economic policy 1034, 151,
25860
main entry 18691
and social conict 323, 21819,
3256, 373
and stagation 3227
Weintraub on 218, 3336, 373
see also monetarism; Phillips curve;
Quantity Theory; stagation;
wages policy
Inexible prices and wages 2759,
2878, 2923
Innovation 67, 125, 178, 1916, 209,
213
see also technical progress
Insideroutsider theory 11516, 292
Institutionalism 68, 145, 196200, 217,
283, 355, 3812
400
Subject index
Institutions 12, 6, 13, 196200,

3514, 38084
Instrumentalism 387
Internal labour markets 382
Interest, theory of
see liquidity preference; money; rate
of interest
International economics 3034, 1315,
16570, 200205, 3469
see also balance-of-paymentsconstrained growth; Bretton
Woods; exchange rates; free trade;
globalization
International Monetary Clearing
Union (IMCU) 204
International Monetary Fund 169,
339, 3489
International monetary system 3034,
1315, 260
Introduction to the Theory of
Employment, by Joan Robinson
211
Investment
and competition 66
and development nance 88
and economic policy 102, 150
and expectations 1368
and nance 88, 144
and growth 93
and interest rates 150, 2289, 242,
246, 301
Kaldor on 173, 183, 3067
Kalecki on 36, 173, 183, 2289
Keynes on 48, 136, 1623, 235,
23940, 246, 301, 307, 372, 376
main entry 205210
Minsky on 1368, 2556
neoclassical theory of 2556
and prots 2948
and saving 1078, 242, 3067, 3723
and uncertainty 21718
ISLM model 259, 265, 313
Italy 60
Japan 18, 19
Joan Robinsons economics 21115
Job guarantee see Employer of last

resort
Journal of Economic Issues 334
96, 21520
Kaleckian economics 2269
Keynesian cross diagram 257
Keyness General Theory 22937
see also General Theory of
Employment, Interest and Money
Keyness Treatise on Money 23741
see also Treatise on Money
Labour market
and employment 11217
Keynes on 2545, 31920
in neoclassical economics 2767
and stagation 326
and unemployment 3748
and wages 38084
see also employment; full
employment; trade unions;
unemployment; wages
Labour theory of value 56, 212
Lender of last resort 24, 579, 146
Lenders risk 78, 1389, 207
Lexicographic preferences 70
Liability management 2021
and banks 21, 234
Davidson on 369
and eective demand 108, 31213
Keynes on 2334, 265, 31213
main entry 2428
and rate of interest 184
Liquidity trap 313
Loanable funds theory of interest 22,
29, 148, 162, 2423, 245, 376
Long-period analysis 445, 47, 112,
212, 214, 32021
Long waves 38, 41
Lucas critique 253, 256
Marginal productivity theory 56, 63,
11314, 1812, 185, 213, 2967,
321
Subject index
Marginalism 24952, 300, 384
Market socialism 31516
Mark-up pricing 61, 63, 67, 92, 120,
168, 183, 188, 2013, 218, 227,
2858, 291, 33032, 334
MarshallLerner condition 17
Marshall Plan 30
Marshallian economics 445, 31920,
385
Marxian economics 923, 154, 182,
184, 212, 228, 289, 3712
Materialism 34
Mathematical methods in economics
364
Megacorp 168
Menu costs 278
Methodological individualism 2, 6, 8,
61
Methodology 467, 96101, 112,
12631, 159, 2523, 271, 321, 387
Microfoundations 61, 74, 106, 10910,
234, 2527
Monetarism 1567, 187, 189, 259, 271,
275
see also Quantity Theory
Monetary circuit, theory of 615, 119
Monetary policy
and banks 21
Bastard Keynesians and 26, 28
and endogenous money 11721, 265
and scal policy 150
and ination 187, 333
Kaldor on 172
Keynes on 240, 246
main entry 25761
neoclassical economics on 187,
300302
in New Classical economics 2715
and rate of interest 300302, 321
see also central banks; ination;
interest rates; Quantity Theory;
Taylor rule
Monetary theory of production 60,
88, 136, 179, 216, 233, 238, 262,
264, 31112, 344
Money
Cambridge economists on 4450
circuit theory of 6064, 267
Davidson on 57
and equilibrium 1289
401
and nance motive 1424

Kaldor on 589
Keynes on 1424, 198, 2336,
23741, 2427, 31112, 3435
Lavoie on 2467
main entry 2615
Moore on 59, 11721, 2467
and Says Law 31012
Wray on 247, 2615
see also endogenous money;
liquidity preference; monetary
policy; neutrality of money; rate
of interest
Monopolistic competition see degree
of monopoly; imperfect
competition; oligopoly
Monopoly 47, 92, 2968
imperfect competition; oligopoly
Monopoly capitalism 228, 371
Multiplier 1719, 35, 48, 144, 230,
26570, 304, 3067, 388
NAIRU (Non-accelerating ination
rate of unemployment) 1567,
279, 324
see also natural rate of
unemployment
National debt 36
Natural rate of interest 187, 190, 240,
301
Natural rate of unemployment 103,
113, 114, 156, 190, 272, 324, 380
see also labour market;
unemployment
Neoclassical economics
Austrian economics and 4
and capital theory 516
and consumer theory 737
and economic methodology 967
and environmental economics 1213
and scal policy 149
and growth theory 92, 94, 176, 223
and income distribution 1812, 185
Joan Robinson on 25
and labour markets 38081, 383
and microfoundations 2526
and production 289, 292
and the rate of interest 300301
402
Subject index
Neoclassical economics (continued)

and Says Law 31213
Sraa on 47
and taxation 328, 331
and time in economic theory 3413
and uncertainty 364
see also general equilibrium theory;
ISLM model; marginalism;
neoclassical synthesis; Walrasian
economics
Neoclassical synthesis 259, 292, 301,
313, 320
Neo-Keynesian economics 157, 173
Neoliberalism 31, 165, 308, 33741
Neo-Ricardian economics 55, 84, 93,
1223, 174, 182, 293
see also Sraan economics
Neutrality of money 6, 19, 63, 186,
233, 236, 238, 282, 284, 3212, 369
New Classical economics 29, 233,
2523, 2713, 2767, 283, 381
New Keynesian economics 29, 77, 87,
89, 145, 230, 233, 256, 275,
27580, 283, 292, 380
New mercantilism 2012, 214
Non-equilibrium 12631, 3559,
388
Non-ergodicity 12, 8, 72, 97, 140,
15960, 21718, 2337, 254, 256,
28084, 3689
Normal cost 2856, 299
Normal prots 2967, 300302
Oligopoly 19, 657, 209, 291, 382
imperfect competition; mark-up
pricing
Open-system thinking 1214, 824,
967, 1678, 193, 2023, 252, 283
see also critical realism
Organicism 14, 967, 128, 162
Outsourcing 1689
Paradox of costs 173
Paradox of debt 22
Paradox of thrift 173, 306
Pasinetti theorem 171
Path dependency 3556, 359, 388
see also history versus equilibrium;
hysteresis
Perfect competition 657, 168, 181,

183
Permanent income hypothesis 73
Phillips curve 27, 42, 1567, 18990,
272, 31819, 3225, 334
Philosophy 1115, 46, 826, 967,
99100, 160, 162, 36064
see also methodology
Planning 31316, 35052
Pluralism 14
Poland 2267
Political business cycle 41
Policy see economic policy
Power 183
Predator-prey models 178
Pre-Keynesian economics 156, 292,
306
see also neoclassical economics
Pricing and prices 19, 467, 657,
1234, 168, 227, 250, 2757,
28591, 299, 3512
see also ination; mark-up pricing
Principle of increasing risk 1467
Principles of Economics, by Alfred
Marshall 249
Principles of Economics, by David
Ricardo 212, 249
Privatization 3524
Probability 16061, 280, 36064,
36670
Procedural rationality 68
Product cycle theory 204
Production 613, 1226, 1678,
28993
Production functions 51, 54, 1812,
213, 2967
see also capital theory
Production of Commodities By Means
Of Commodities, by Piero Sraa
47, 50, 171, 1734, 21314
see also Sraan economics
Productivity growth 66, 174, 2245,
375
Prot, rate of see rate of prot
Prots
and budget decits 356
in circuit theory 63
in classical political economy 319
Kaldor on 1713, 1834, 2234
Kalecki on 356, 2267, 32930
Subject index
Keynes on 239
main entry 2948
and taxation 32930
see also rate of prot
Propensity to consume 726, 107,
1835, 197, 237, 329, 371
Propensity to save 246, 371
Pseudo-production function 213
Psychology and economics 1323,
2234
Public employment 37, 102, 1578,
3789
Public works 35
Quantity Theory of money 45, 48, 58,
117, 1867, 238, 242, 246, 333
see also monetarism
Quarterly Journal of Economics 159,
171
Radclie Committee 172
Rate of interest
in circuit theory 64
in classical economics 31011,
376
and credit rationing 81
and scal policy 150
Friedman on 2712
Kaldor on 589
Kalecki on 589
Keynes on 108, 184, 2259, 240,
2427, 265, 307, 312, 376
main entry 299303
and monetary policy 25760,
265
Moore on 212, 589
in neoclassical theory 212, 12021,
124, 187, 31011, 376
Pasinetti on 184
and prots 295
Sraa on 321
see also credit rationing; interest;
ISLM model; liquidity
preference; liquidity trap
Rate of prot 173, 177, 1845, 21213,
293, 297, 300, 302, 319
Rational expectations 103, 132, 233,
2534, 258, 273, 28081, 284, 307,
387
see also New Classical Economics
403
Rationalism 8, 11, 363

Rationality 68, 2715, 276, 282, 3423,
360, 3623, 366
Realism 2, 6, 826
see also critical realism
Rent 249
Reswitching of techniques 523, 55,
213
Rhetoric 98
Ricardian equivalence 2723
Risk 78, 1389, 207, 296, 366
see also principle of increasing risk;
uncertainty
Russia 169, 31415, 351
Saving
and development nance 88
and eective demand 108, 2346,
246, 269, 312
and investment 93, 108, 23940,
312, 372, 376
Keynes on 2346, 23940, 242, 246,
312, 376
main entry 3049
Says Law
Bastard Keynesianism and 29
and full employment 1345, 154
Keynes on 109, 162, 23033, 376
main entry 30913
and neoclassical theory 1312
New Classical economics and 272
New Keynesian economics and 29
and saving 304
Segmented labour markets 3824
Shock therapy 352
Short-period analysis 448
Social democracy 33741, 373
Socialism 227, 229, 31318, 337,
35054, 358
see also Marxian economics
Soviet Union 31415, 337, 341, 373
Speculation 3032, 90, 1389, 143,
146, 168, 202, 207, 222, 2445,
346, 349
Speculative demand for money 143,
2436
404
Subject index
Sraan economics 556, 72, 216,

26970, 31822, 356, 359
see also classical political economy;
neo-Ricardian economics
Stagation 2, 7, 216, 3227, 332,
see also ination; unemployment
Stagnation 209, 228, 3723
State theory of money 2623
Structural change 934, 178, 3778
Subcontracting 168
Subjectivism 160, 182
see also Austrian economics
Substitution 69, 71, 104, 2336
Supply-side economics 331
Surplus, economic 4950, 153, 270,
295, 31920, 371
see also Marxian economics; Sraan
economics
Sustainability 1256
Target rate of return pricing 286
Taxation 172, 2223, 2623, 317,
32832
see also budget decits; scal policy;
tax-based incomes policy; Tobin
tax
Tax-based incomes policy (TIP) 103,
151, 188, 3328
see also wages policy
Taylor rule of monetary policy 121,
187, 259
Technical progress 54, 95, 1767, 213,
3757
see also innovation
Technical progress function 224
Technology gap 2034
Temporary equilibrium 3856
Third Way 33741
Thirlwalls law 16, 18, 94, 217
Time in economic theory 2, 6, 445,
108, 1279, 163, 214, 284, 3416
see also expectations; history and
equilibrium; hysteresis;
uncertainty
Time-dependence see hysteresis
Tobin tax 103, 3469
Trade, theory of 167, 200205
Trade cycles see business cycles
Trade unions 11516, 185, 188, 277,
373, 376, 380, 382
Transition economies 35054

Transnational corporations 168
Traverse 3559
see also history and equilibrium;
time in economic theory
Treasury view 304
Treatise on Money, by J.M. Keynes 35,
47, 54, 60, 79, 88, 2056, 211,
23741, 247, 261
Treatise on Probability, by J.M. Keynes
160, 1623, 35965
Trieste summer school 320
Twin decits 14951
Uncertainty
in Austrian economics 69, 13
and credit rationing 77
Davidson on 69, 13, 280, 342
and environmental economics 1245
and equilibrium 128
and exchange rates 133
and investment 207, 209, 21718,
290, 2956
Kalecki on 228
Keynes on 15964, 207, 2345, 271,
280, 31213, 3434, 364
main entry 36670
and New Classical economics 274
and neoclassical economics 254,
256, 321
Sraa on 321
and Walrasian economics 388
see also expectations; principle of
increasing risk
Underconsumption 37074
Unemployment
and competition 65
Davidson on 198
economic policy and 1034, 1556,
259
and ination 41, 103, 156, 3227
Keynes on 478, 1623, 230, 234,
241, 265
main entry 3749
Marx on 154
monetary policy and 159
New Keynesian economics on 29
wage rigidity and 113
Subject index
Walrasian economics and 386, 388
employment; NAIRU; natural
rate of unemployment; Phillips
curve; stagation
United Nations Development
Programme (UNDP) 346, 348
Velocity of circulation of money 272,
334
Verdoorns Law 224
Wage dierentials 383
Wage-cost mark-up 188, 218, 3346
employment
Wages, money
and employment 185, 250, 2923,
373
and ination 1889, 3256, 3345
main entry 38084
in New Keynesian economics 2757
and real wages 250
rigidity of 113, 2757
see also labour market; Phillips
curve; Pigou eect; tax-based
incomes policy; trade unions;
wages policy
Wages, real
conict over 189, 302, 3256, 3734
405
and eective demand 2545, 373

and employment 109, 11315,
1812, 185, 311, 374
and growth 173, 185
Kalecki on 331, 3734
Keynes on 2545
main entry 38084
Marx on 377
and money wages 250
and the rate of interest 300, 302
Ricardo on 297
Sraa on 319
see also eciency wages; marginal
productivity
Wages policy 103, 151, 157, 188,
3327
see also ination; tax-based incomes
policy
Walrasian economics 26, 342, 345,
381, 3848
see also general equilibrium theory;
ISLM model; neoclassical
economics; neoclassical synthesis
Washington consensus 33741
Weight of argument 161, 163
Welfare economics 1226
Welfare state 33741, 354
World Bank 348
World Trade Organisation 341
Yugoslavia 316

The Elgar Companion To Post Keynesian Economics

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Elgar Companion To Post Keynesian Economics

Uploaded by

Copyright:

Available Formats

THE ELGAR COMPANION TO

POST KEYNESIAN ECONOMICS

The Elgar Companion to

J.E. King 2003

A catalogue record for this book

Library of Congress Cataloguing in Publication Data

ISBN 1 84064 630 6 (cased)

Fiscal Policy J.W. Nevile

Traverse Peter Kriesler

Conceptualizations of economic processes

Harcourt, G.C.: Jesus College, Cambridge, UK

ODonnell, Rod: Macquarie University, Sydney, Australia

Webster, Elizabeth: University of Melbourne, Melbourne, Australia

element of truth: one way of appreciating what Post Keynesians do believe is

relationship between an agent and the social structure? Two traditional

Austrian school of economics

Austrian School of Economics

Austrian school of economics

Austrian school of economics

As Vaughn (1994) acknowledges, there are two strands in contemporary

Austrian school of economics

analysis, in particular to adoption of aggregate concepts. This objection

Austrian school of economics

Austrian school of economics

Babylonian Mode of Thought

Babylonian mode of thought

representation of what he called the Babylonian tradition in mathematics,

Babylonian mode of thought

Babylonian mode of thought

the economist understands the economic process in terms of production or

Balance-of-payments-constrained economic growth 15

Balance-of-payments-constrained Economic Growth

Balance-of-payments-constrained economic growth

Balance-of-payments-constrained economic growth 17

Balance-of-payments-constrained economic growth

obtained using the expression for Thirlwalls law, yB z/. (Alternatively,

Balance-of-payments-constrained economic growth 19

banking system. Liability management refers to the ability of banks to vary

semantic by pointing out that the horizontalist position is overstated in

hostile in their reviews. The doyen of American economics, Paul Samuelson

Paul Samuelsons response to the General Theory is instructive. While he

moving through historical time is essential to the message of Keynes.

ties, often became disequilibrating instead of equilibrating, or instead of

sector to the industrial capitalist and working classes, leading to an increase

in capitalist economies, and he favoured public works in order to increase

deficit increases profits simultaneously with output and employment.

Cambridge Economic Tradition

Cambridge economic tradition 45

Cambridge economic tradition

Cambridge economic tradition 47

Cambridge economic tradition

Cambridge economic tradition 49

Cambridge economic tradition

was integrated into the Cambridge approach and tradition. Pasinettis

Sraa, P. (1960), Production of Commodities by Means of Commodities. Prelude to a Critique

One way of interpreting the Cambridge controversy is in terms of

This is fine as far as it goes as long, as previously mentioned, as there is

as suggested by reference to the marginal product of capital for the as if

inputoutput matrices and the wageprofit trade-o they sustain

Keynes himself believed that central banks, besides performing their

The monetary circuit

In the theory of the monetary circuit there is a strict distinction between

firms cannot operate under the price competition of those commodity